Welsh Rates of Income Tax HMRC Annual Report 2026
Published 22 September 2026
Applies to Wales
© Crown copyright 2026
This publication is licensed under the terms of the Open Government Licence v3.0 except where otherwise stated. To view this licence, visit nationalarchives.gov.uk/doc/open-government-licence/version/3 or write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email: psi@nationalarchives.gov.uk.
Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned.
This publication is available at https://www.gov.uk/government/publications/welsh-rates-of-income-tax-hmrc-annual-report-2026/welsh-rates-of-income-tax-hmrc-annual-report-2026
Welsh Rates of Income Tax at a glance
Our work in numbers
- 1,660,400 Welsh taxpayers identified as at 2024 to 2025
- £3,300 million Welsh Rates of Income Tax outturn attributable to the Welsh Government budget
- 2 Welsh Rates of Income Tax (WRIT) Board meetings held
- 9,855 letters sent as part of the Welsh third-party data assurance exercise
- 1 pieces of data analysis published
- 665 invalid or blank postcodes updated
- 457 new postcodes updated based on Office for National Statistics (ONS) data
See Annex D: Metrics dictionary for the sources, scope, and caveats of these figures.
What changed in 2025 to 2026
The Welsh Government set the Welsh Rates of Income Tax at 10p in each band. This maintained parity with Income Tax rates for taxpayers in England and Northern Ireland.
In the 2025 Budget, the UK Government announced a measure to introduce separate tax rates for property income from April 2027. The UK Government is providing the Welsh Government with the ability to set Property Income rates in line with their current Income Tax powers in the fiscal framework.
The Service Level Agreement was updated to set out that in the absence of a Welsh Rate Resolution, the Government of Wales Act 2006 enables the UK Government to lay a Treasury Order to allow the continued operation of PAYE for Welsh taxpayers at the rates and bands of the previous tax year.
HMRC published an updated analysis of long-term trends in intra-UK migration between Scotland, Wales, and the rest of the UK. See Intra-UK migration of taxpayers for the full analysis and results.
What is changing in 2026 to 2027
The Welsh Rates of Income Tax are being maintained at 10p in each band. Full details of the tax bands and rates in 2026 to 2027 are available on GOV.UK.
The Service Level Agreement has been updated to incorporate the Memorandum of Understanding.
HMRC and the Welsh Government will continue to prepare for the introduction of separate tax rates for property income.
Key messages
- most Welsh taxpayers experience WRIT through HMRC’s standard services and notice the same level of service as other Income Tax payers in the UK. See Taxpayer Experience
- the 2024 to 2025, the final WRIT outturn attributable to the Welsh Government Budget was £3,300 million. See 2024 to 2025 Outturn Calculation
- HMRC will stop producing the provisional estimate for efficiency. See Provisional Estimate
- the latest (2025) third-party data assurance exercise found that HMRC residency information is between 98 to 99% accurate. The next exercise will be held in 2027. See Key Controls to Assess and Collect Income Tax and Annex A Findings from Third Party Data Assurance Exercise
- HMRC assesses the risk of WRIT non-compliance in 2025 to 2026 and 2026 to 2027 as low. See Assessing and Mitigating the Risk of WRIT Non-Compliance
- HMRC recharged the Welsh Government £0.43 million. See Cost of Administering WRIT
Governance
HMRC and the Welsh Government oversee the operation of WRIT through formal governance structures. The WRIT Board provides senior oversight of performance, delivery, costs, and key risks.
Taxpayer experience
Conclusion
In most cases, Welsh taxpayers experience WRIT through HMRC’s standard PAYE and Self Assessment services and notice little difference in how HMRC interacts with them compared with taxpayers elsewhere in the UK.
HMRC’s objective is to provide the same level of customer service, support and transparency to Welsh taxpayers as is provided to other Income Tax taxpayers, while ensuring that guidance, systems and communications reflect the Welsh rates and thresholds that applied in 2025 to 2026.
Evidence
Between April 2025 to July 2025, there were 6 calls to the Welsh Income tax telephone route within the HMRC Personal Taxes helpline. This gives customers generic pre-recorded WRIT messages prior to speaking to a HMRC customer advisor. The average call wait time was 09:04 minutes.
HMRC also monitored complaints. HMRC tracks all Welsh income tax customer complaints to ensure they are processed within the HMRC customer service targets. HMRC received 0 complaints from Welsh Income Tax customers about the Welsh Rates of Income Tax.
During 2025 to 2026, HMRC reviewed and updated WRIT guidance products ahead of the start of the tax year so that information available to Welsh taxpayers on GOV.UK remained accurate and commensurate with that available to taxpayers in the rest of the UK. HMRC also updated calculators, systems, and tax tables used in administering Welsh rates and thresholds, including tools used by employers and compliance caseworkers, to support accurate operation of the 2025 to 2026 Welsh rates.
HMRC also translated key documents into Welsh and published the English and Welsh versions at the same time. This includes the Service Level Agreement and the Annual Report 2025.
Limitations
The available evidence is drawn mainly from operational indicators and business intelligence rather than a standalone end-to-end measure of the Welsh taxpayer experience. As a result, HMRC can identify the main points of contact and common issues, but the evidence base is stronger in some parts of the customer journey than others.
The customer complaints data does not represent all the customer complaints HMRC receives from Welsh taxpayers. Tier 1 complaints arrive and are responded to by the business area.
The customer call data does not represent all the telephone calls HMRC receives from Welsh taxpayers as they may use different phone lines. Devolved telephony data is unavailable for the majority of the tax year due to changed reporting information. The figures above are based on calls between April 2025 to July 2025.
Future work
HMRC will continue to use insight from the business intelligence reporting and engagement with the Welsh Government to identify further opportunities to improve the experience of Welsh taxpayers. This includes continuing to refine guidance, communications and service design where evidence suggests that Welsh taxpayers may benefit from clearer information or smoother processes.
Assurance
2024 to 2025 Outturn calculation
Conclusion
For 2024 to 2025, the total amount of Non-Saving/Non-Dividend Income Tax generated by Welsh taxpayers was £7,765 million. The final Welsh Rates of Income Tax outturn attributable to the Welsh Government budget was £3,300 million.
Evidence
HMRC calculates the outturn using available administrative data and publishes a full statistical release. The full report is available at Welsh Income Tax Outturn Statistics: 2024 to 2025.
The outturn figure was calculated using actual data available from HMRC’s Pay As You Earn and Self Assessment systems, about 3 months after the Self Assessment filing peak. Further detail is provided in the WRIT Outturn Statistics 2024 to 2025, which includes the information shown in the HMRC accounts and additional breakdowns for Welsh taxpayers and taxpayers in the rest of the UK.
Limitations
A small component (under 5%) of the outturn figure is estimated, drawing on historic data such as filing patterns and compliance activity. This reflects trade-offs between timeliness and budgetary certainty. A degree of estimation will always be required because some tax liabilities continue to be established for several years after the end of the tax year.
Future work
The 2025 to 2026 outturn figure will be published in 2027.
Provisional Estimate
Conclusion
HMRC and the Welsh Government have agreed to stop producing the provisional estimate. The National Audit Office has been informed of this decision.
Evidence
In previous years, HMRC calculated a provisional estimate of WRIT revenue expected after the end of a tax year. This was published in the Annual Report and Accounts but served no budgetary purpose and was superseded by the Office for Budget Responsibility’s WRIT forecast., This is produced by HMRC using the same method and is used in both the Welsh Government and UK Government Budgets.
Limitations
The provisional estimate duplicated analytical effort because it used the same methodology as the subsequent Office for Budget Responsibility forecast, which is used for budgetary purposes.
Future work
The analytical resource usually assigned to the provisional estimate will be redirected to other high-priority work funded by the Welsh Government.
Key controls to assess and collect Income Tax
Conclusion
HMRC has a high degree of confidence in the accuracy of its Welsh taxpayer identification and postcode records. HMRC’s residency information is between 98 to 99% accurate, postcode records show a high level of accuracy, and ongoing assurance activity helps protect revenue for the Welsh Government.
Evidence
HMRC engages an external supplier to carry out a third-party data assurance exercise to assess the accuracy of its identification of Welsh taxpayers. The latest report is available in Annex A, Third-Party Data Assurance Exercise 2025 and shows that HMRC’s residency information is between 98 to 99% accurate.
HMRC also conducts address assurance scans to identify and correct postcode data quality issues, including invalid postcodes and blank postcodes containing indicators that an address is in Wales. In 2025 to 2026, HMRC corrected 665 invalid or blank postcodes in Wales.
HMRC receives quarterly postcode updates from the ONS and updates its processes for flagging Welsh residency to reflect changes including new build properties and the subdivision of existing properties. In 2025 to 2026, HMRC received and updated 457 new postcodes in Wales based on information from the ONS.
Finally, HMRC have set a flag to review all residential moves into one of the Welsh Cross Border postcodes. HMRC review these moves to ensure the correct residency indicator is set. There were 382 Welsh cross border moves reviewed in 2025 to 2026 and of those, 75 resulted in amending the customer’s residency status.
Limitations
There is no definitive data set of Welsh residents against which to judge success, so assurance relies on corroborative exercises rather than direct verification.
HMRC receives address data from a variety of sources, including individuals, employers, and other government departments, which means that while most records are accurate, a small minority will always require correction.
Future work
The next third-party data assurance exercise will be held in 2027.
The next address assurance scan will be completed in 2026 to 2027, and HMRC will continue to receive and apply quarterly postcode updates from the ONS in 2026 to 2027.
Assessing and mitigating the risk of WRIT non-compliance
Conclusion
WRIT non-compliance refers to individuals who deliberately manipulate their residency information to avoid paying WRIT. This could be achieved by:
- an individual changing the address held by HMRC
- an individual with 2 homes claiming to spend most of their time where tax rates are most favourable
- an individual neglecting to update their address with HMRC
HMRC assesses the risk of WRIT non-compliance as low.
Evidence
HMRC’s assessment is informed by the differential between Welsh and UK Income Tax rates and the evidence gathered from previous activity and is overseen by the Welsh Income Tax Board.
In 2025 to 2026, the WRIT matched the UK rates so neither set of rates was more favourable and there was no motivation for WRIT non-compliance. Therefore, HMRC assessed the risk of WRIT non-compliance as low with no evidence of widespread behavioural change. Given the 2026 to 2027 rates remained the same, HMRC’s initial view is that the risk will remain low in 2026 to 2027. HMRC’s assessment was reviewed by the Welsh Income Tax Board.
Recent analytical work has examined a range of indicators relevant to compliance risk, including:
- migration patterns to and from Wales which can impact taxpayer identification
- employer operation of tax codes
HMRC’s address assurance scans show no evidence that individuals living in Wales are manipulating their addresses to claim to live elsewhere. If this were taking place on a significant scale, HMRC would likely see an increase in the number of Welsh taxpayers becoming UK taxpayers. Similarly, the third-party data assurance exercise gives HMRC confidence that the correct residency status has been applied in 98 to 99% of cases, with the remaining 1 to 2% not necessarily wrong, but uncorroborated.
HMRC’s longitudinal analysis of intra-UK migration found that there was a substantial increase in net migration to Wales during the COVID-19 pandemic.
Limitations
The longitudinal migration analysis does not provide specific insight into non-compliance. However, it does provide useful insight into the potential motivation for taxpayers to undertake behavioural change in response to diverging tax rates, which is relevant to our understanding of the risk of WRIT non-compliance.
HMRC cannot draw conclusions from the longitudinal migration analysis because the level of migration without Income Tax devolution is unknown.
In addition, it’s possible that a small number of cases of WRIT non-compliance may not be identified, and any activity designed to find such cases would be unlikely to be cost effective even if it recovered yield.
Future Work:
In 2026 to 2027, HMRC will continue to monitor WRIT and keep its assessment of WRIT non-compliance risk under review through its established governance arrangements. HMRC will also continue its analytical work on migration patterns, complex residency cases, employer coding issues, and wider behavioural indicators.
Cost of Administering WRIT
Conclusion:
In 2025 to 2026, HMRC recharged the Welsh Government £0.43 million as the net additional cost wholly and necessarily incurred as a result of the implementation and administration of WRIT. HMRC has a robust process in place to identify the costs of administering WRIT accurately.
Evidence:
HMRC provides the Welsh Government with a report each month outlining the quarterly costs to be recharged. Once the WRIT Board is satisfied that the costs are accurate, HMRC raises an invoice, supported by evidence that allows the Welsh Government to assure itself of the accuracy of the costs to be recharged. Further detail is set out in the Rechargeable Costs Framework annexed to the Service Level Agreement.
| Quarter | Implementation and Operating Costs Recharged in 2025 to 2026 (£ million) |
|---|---|
| Quarter 1 | £0.08 |
| Quarter 2 | £0.13 |
| Quarter 3 | £0.13 |
| Quarter 4 | £0.09 |
| Full year | £0.43 |
| Cost (£ million) | 2021-22 | 2022-23 | 2023-24 | 2024-25 | 2025-2026 |
|---|---|---|---|---|---|
| Operating Costs | £0.44 | £0.42 | £0.43 | £0.38 | £0.43 |
| Total Cost | £0.44 | £0.42 | £0.43 | £0.38 | £0.43 |
Limitations:
Comparisons across years should be read with care.
Future Work:
HMRC will continue to provide monthly cost reports, maintain WRIT Board oversight of rechargeable costs, and support the Welsh Government with evidence to assure future recharging.
SLA Mapping Index
This index maps the main responsibilities and performance requirements set out in the Service Level Agreement for operation of WRIT by HMRC, as described on GOV.UK, to the sections of this report and annexes where supporting evidence is provided.
| SLA Requirement | Expectations of HMRC | Evidence in this Report and Annexes |
|---|---|---|
| Taxpayer identification and residency controls | HMRC should operate processes and controls that support accurate identification of Welsh taxpayers and correct application of residency status. | Key Controls to Assess and Collect Income Tax; Annex A; Annex B. |
| Collection and management of WRIT revenues | HMRC remains accountable for the collection and management of WRIT revenues and must evidence the correct amounts brought to account. | Key Messages; 24-25 Outturn Calculation; Key Controls to Assess and Collect Income Tax. |
| Service quality for Welsh taxpayers | HMRC should provide a consistent quality of service to Welsh taxpayers, equivalent to that provided to taxpayers elsewhere in the UK. | Welsh Rate of Income Tax at a Glance; What Changed in 2025 to 2026; Key Messages; Taxpayer Experience. |
| Data and analysis for forecasting, policy, and budgeting | The Welsh Government needs sufficient data and analysis from HMRC to support forecasting, policy development, rate setting and budgeting, and to manage forecast variance. | Welsh Rate of Income Tax at a Glance; What Changed in 2025 to 2026; 24-25 Outturn Calculation; Provisional Estimate; Annex C. |
| Compliance risk management | HMRC should assess and manage the risk of WRIT non-compliance and keep the risk under review. | Key Messages; Assessing and Mitigating the Risk of WRIT Non-Compliance; Annex B. |
| Annual reporting and transparency | HMRC should evidence that it is operating WRIT efficiently and effectively and delivering the agreed services. | Welsh Rate of Income Tax at a Glance; Taxpayer Experience; Assurance; Annex. |
| Operation of PAYE, Self Assessment and supporting systems | HMRC should maintain systems, processes, tax codes, guidance and operational arrangements needed to administer WRIT within the UK tax system. | What Changed in 2025 to 2026; Taxpayer Experience; Key Controls to Assess and Collect Income Tax. |
| Governance, monitoring and joint working | HMRC and the Welsh Government should maintain governance, monitoring and reporting arrangements to oversee the operation of WRIT. | Welsh Rate of Income Tax at a Glance; Governance; Assessing and Mitigating the Risk of WRIT Non-Compliance; Cost of Administering WRIT. |
| Rechargeable costs and value for money | The Welsh Government must meet net additional costs incurred by HMRC, and HMRC should evidence those costs fairly and accurately. | Cost of Administering WRIT. |
| Continuous improvement | HMRC is committed to continuous improvement in its operation of WRIT, where improvements are possible and cost effective. | What Changed in 2025 to 2026; Taxpayer Experience; Future Work subsections across Assurance. |
Annex A: Findings from Third Party Data Assurance Exercise
Outcome of the Third-Party Data Clash matching exercise
Table 1a: GB Group unmatched records removed by KAI
| GB Group unmatched records | Number removed by KAI |
|---|---|
| Duplicates | 2,622 |
| Deceased | 1,788 |
| Under 18 | 66,117 |
| Resident abroad | 7,463 |
Table 1b: GB Group unmatched records matched by KAI
| GB Group unmatched records | Number matched by KAI |
|---|---|
| Matched to Wales | 581,438 |
| Matched to rUK | 9,022 |
| Unmatched | 106,425 |
Table 1c: GB Group unmatched records totals matched by GBG and KAI and percentage adjusted totals
| GB Group unmatched records | Total matched by GBG and KAI | % adjusted total |
|---|---|---|
| Matched to Wales | 2,901,183 | 96.03% |
| Matched to rUK | 13,375 | 0.44% |
| Unmatched | 106,425 | 3.52% |
Note: Where ‘adjusted total’ refers to the initial number of Welsh individuals with duplicates, deceased, those aged under 18 and those resident abroad removed, which is 3,020,983 in this instance.
Methodology
KAI sent GB Group approximately 91 million records sourced from CID (Citizen Identification Framework), as part of assurance of HMRC address data.
In order to provide assurance of HMRC address data, GB Group performed a matching exercise of the CID records against third party data sources, such as the electoral register and credit records.
Following the completion of this exercise, GB Group sent KAI reports outlining the number of records successfully matched to a country, the number of contradictory address records (by country) and the number of records that are not matched to any address in third party data.
Alongside this report, GB Group sent KAI a list of NINOs (National Insurance numbers) and TRNs (Temporary Reference numbers) [footnote 1] that did not match to a third-party source.
KAI seek to corroborate these records and identify their most recent address/country information in HMRC data. This is performed by:
- removing duplicate NINO’s and TRN’s from the initial list
- matching the records to a number of HMRC systems, meaning CESA/NPS and RTI to identify: deceased individuals, individuals under 18 (at the time the GB Group matching exercise was performed) and individuals residing abroad
- records which are identified as deceased, under 18 or residing abroad are removed from the initial unmatched list
- remaining records that match to HMRC systems are assigned to a region/country based upon a valid postcode and the corresponding GOR (region)/ITL2 region associated with the postcode. This is obtained by matching extracted postcodes to a postcode lookup table for GOR Codes and assigning a corresponding Income Tax Regime region based on the first letter of this variable
Results – GB Group Matching Exercise
Of 3,098,973 Welsh records:
- 2,319,745 (74.86%) records were matched to a Welsh address that had a Welsh address in CID
- 4,353 (0.14%) records were matched to an rUK address but had a Welsh address in CID
- 774,875 (25.00%) Welsh records were not matched to any address
For the 4,353 records that were matched to an rUK address, a letter was sent to the individual asking them to update their address data.
GB Group also found 5,502 records in the CID data with an rUK address but matched to a Welsh address in the third-party data – these individuals were also sent a letter asking them to update their address. For this exercise, we have treated these cases as Welsh residence.
Results – KAI Matching Exercise
As outlined, KAI matched the 774,875 uncorroborated records to HMRC systems and:
- removed 2,622 duplicate NINO’s/TRN’s from the initial list
- removed 66,117 records where the individual was under the age of 18 when the assurance exercise was carried out:
- of these individuals: 39 were under 16, 32,443 were 16 and 33,635 were 17
- removed 7,463 records where the individual was identified as residing abroad
- removed 1,788 records where the individual was identified as deceased
- identified 581,438 records where the individual had a Welsh GOR Code (83.43% of the initial, adjusted unmatched list - 696,885)
- identified 9,022 records where the individual had an rUK GOR Code
- we were unable to corroborate 106,425 records. Of these:
- 102,837 (96.63%) were TRNs which is 3.40% of the adjusted list of 3,020,983 records sent to GB Group (with duplicates, deceased, under 18 and overseas cases excluded)
- 3,588 (3.37%) were NINOs of which 2,863 have no postcode information associated to them. Another 720 have invalid postcodes therefore we were unable to obtain a GOR_CODE
See table 1 for further detail.
Conclusion
After the completion of both KAI’s and GB Groups’ components of this third-party matching exercise, it is shown that HMRC’s taxpayer status was correct in 96.03% of cases (when duplicates, deceased and overseas cases are excluded). However, when considering cases where a match was made, 2,901,183 out of 2,914,558 cases were matched as Welsh (99.54% of this adjusted total).
Therefore, across the 2 stages, only 0.46% of these records were identified as residing in rUK when CID provided a Welsh address. And just 106,425 (3.43%) of the initial list in CID remain uncorroborated, with most (96.63%) of these being TRNs.
When considering this figure in the context of the overall accuracy of HMRC’s identification of Welsh taxpayers it is important to bear in mind that this population does not correspond to distinct, active Welsh taxpayers in any single given year. It represents records (as opposed to individuals) held in HMRC’s CID data, which could be historic accounts that are no longer active; or TRNs which belonged to individuals that have subsequently been allocated a NINO. So, an uncorroborated taxpayer status doesn’t mean an incorrect taxpayer status.
We are therefore confident that the proportion of correctly identified Welsh cases in HMRC’s CID will likely be around 98%-99% when we only consider records which are currently active in SA or RTI. The uncorroborated cases are likely to be in respect of ceased or inactive accounts which explains why KAI have been unable to identify them in either the SA or RTI data. There may also be some cases where postcode information has been incorrectly input, or an expired postcode has been input meaning that a GOR Code cannot be allocated to the record. Therefore, there is no risk to WRIT of failing to corroborate these records as Welsh since there will be no Income Tax liability for them.
This gives confidence that HMRC will have correctly identified around 99% of currently active Welsh taxpayers from our CID data as living in Wales.
Annex B: Scottish and Welsh Income Tax Codes
Background
The Data Quality Team within the Chief Data Office were commissioned by Operational Excellence and Policy to build a Power BI report to provide information on Scottish (‘S’) and Welsh (‘C’) Income Tax codes.
The key aims of the report were to:
- Show the size of the Employers, Pension Administrators and 3rd Parties that are not operating the ‘S’ or ‘C’ codes correctly, meaning is the error made predominantly by small, medium, or large Employers, Pension Administrators and 3rd Parties?
- What sectors are these Employers, Pension administrators and 3rd Parties in that do not operate the ‘S’ or ‘C’ codes correctly?
- Is it the same Employers, Pension administrators and 3rd Parties across the years that incorrectly apply ‘S’ or ‘C’ codes?
- Does the S’ or ‘C’ code error affect high-rate taxpayers or low-rate taxpayers?
- What software product name and version are used, and which ones are causing the most discrepancies?
Upon analysis of the data, further aims were added:
- The Employers, Pension administrators and 3rd Parties who have No Net Liability (NNL) employees and are not operating the ‘S’ or ‘C’ codes correctly.
- Employments that have had a P6/P9 code issued within the last 60 days and the associated Employers, Pension administrators and 3rd Parties, business sizes and products not operating the ‘S’ or ‘C’ codes correctly.
As National Insurance and PAYE System, NPS, is regarded as the master of citizen data for HMRC, when looking at discrepancies within this report, it will be focused on where NPS is showing the employment/payment to be on Scottish or Welsh income tax code and Real Time Information, RTI, to not be showing as Scottish or Welsh.
Financial Years 2023, 2024, 2025, 2026 and year to date, year to date (YTD), 2027 (6 April – 6 June) were used for this analysis. Within this report, when looking at specific scenarios, visuals will be for YTD 2027.
Payment Level
To determine the scale of the problem initially, data was kept at payment level, this means it was showing the tax code each time someone got paid. For example, if someone works at 2 companies and gets paid monthly, they will appear 24 times within a financial year worth of data.
This view was chosen as there is potential that an employee could have the incorrect tax code for a percentage of the year and the correct tax code for the other percentage. For example, someone who is paid monthly could have 11 payments with the correct tax code and 1 month on the incorrect tax code. Results for each financial year are shown in Figure 1 and Figure 2.
Figure 1: Breakdown of the number of payments across the Tax Years and the associated match status (note)
| Match Status | 2023 | 2024 | 2025 | 2026 | 2027 |
|---|---|---|---|---|---|
| Neither Tax Code Present | 689,461,409 | 690,523,874 | 687,341,303 | 579,449,781 | 233,461,200 |
| NPS Scottish/RTI Not | 2,165,544 | 2,156,162 | 2,019,521 | 1,441,533 | 542,889 |
| NPS Welsh/RTI Not | 2,101,090 | 2,057,998 | 2,006,752 | 1,431,654 | 540,820 |
| RTI Scottish/NPS Welsh | 2,594 | 2,761 | 2,874 | 1,463 | 598 |
| RTI Scottish/NPS Not | 240,512 | 213,774 | 208,823 | 136,454 | 60,308 |
| RTI Welsh/NPS Scottish | 2,633 | 2,627 | 3,116 | 1,573 | 608 |
| RTI Welsh/NPS Not | 159,823 | 145,115 | 139,894 | 78,419 | 32,052 |
| Scottish Tax Code Match | 64,233,458 | 64,648,900 | 64,619,770 | 54,435,035 | 22,265,320 |
| Welsh Tax Code Match | 34,428,692 | 34,484,882 | 34,244,918 | 29,348,145 | 11,811,578 |
| Total | 792,795,755 | 794,236,093 | 790,586,971 | 666,324,057 | 268,715,373 |
Note: 2027 YTD (6 April – 19 to 6 August 2026) and Neither Tax Code Present was not used in the remainder of the report visuals. For definition of each match status please see definitions.
Figure 2: Breakdown of the number of payments across the Tax Years and the associated match status (note)
| Match status | Number of payments 2023 (millions) | Number of payments 2024 (millions) | Number of payments 2025 (millions) | Number of payments 2026 (millions) | Number of payments 2027 (millions) |
|---|---|---|---|---|---|
| Scottish Tax Code Match | 64 | 65 | 65 | 63 | 22 |
| Welsh Tax Code Match | 34 | 34 | 34 | 34 | 12 |
| NPS Scottish/RTI Not | 2 | 2 | 2 | 2 | 1 |
| NPS Welsh/RTI Not | 2 | 2 | 2 | 2 | 1 |
| RTI Scottish/NPS Not | 0 | 0 | 0 | 0 | 0 |
| RTI Welsh/NPS Not | 0 | 0 | 0 | 0 | 0 |
| RTI Welsh/NPS Scottish | 0 | 0 | 0 | 0 | 0 |
| RTI Scottish/NPS Welsh | 0 | 0 | 0 | 0 | 0 |
Note: 2027 YTD (6 April 2026 to 6 August 2026). For definition of each match status please see definitions.
Employment level
When looking at the National Insurance Number, NINo, or Temporary Reference Number, TRN, of an individual we refer to this as the employment level. We can’t assume that an individual only has one employment, so we combine the NINo/TRN with an employment sequence number, which shows the order of an individual’s employment, to represent the individual’s employment(s).
For this analysis, we looked at the employment on their latest payment, this is the last time the individual got paid. For 2027, the data was extracted on 19 August, so it was the latest payments on or before the 19 August.
Employers, pension administrators, and third parties
Employers, pension administrators, and third parties were divided into business sizes. Business sizes for this analysis were defined using the Department for Business, Energy and Industrial Strategies definitions, and are defined in Table 1.
| Term | Description |
|---|---|
| Large business | A business with 250 or more employees |
| Medium-sized business | A business with 50 to 249 employees |
| Small business | A business with 10 to 49 employees |
| Micro business | A business with 1 to 9 employees |
In 2027, the number of employers, pension administrators, and third parties is represented in Figure 3. Where an employers, pension administrators, and third parties employment size couldn’t be determined this is represented by blank.
Large businesses represent 74.91% of the employers, pension administrators, and third parties administrating Scottish/Welsh tax codes.
Figure 3: Percentage split of employers, pension administrators, and third parties business sizes administrating Scottish/Welsh tax codes for 2027 YTD (6 April 2026 to 6 August 2026).
| Business size | Percentage |
|---|---|
| Large business | 74.91% |
| Medium-sized business | 8.73% |
| Small business | 9.05% |
| Micro business | 7.31% |
For match category, NPS Scottish/RTI not, shown in Figure 4, microbusinesses represent the greatest number of discrepancies, 34.85%, with large business representing the next, 26.90%. For match category, NPS Welsh/RTI not, shown in Figure 4, micro businesses represent the greatest number of discrepancies, 34.85%, with large businesses representing the next, 26.90%.
Figure 4: Percentage split of employers, pension administrators, and third parties business sizes who shown NPS as Welsh income code and RTI is not for 2027 YTD (6 April 2026 to 6 August 2026).
| Business size | Percentage |
|---|---|
| Large business | 26.90% |
| Medium-sized business | 14.01% |
| Small business | 24.21% |
| Micro business | 34.85% |
Sectors
UK Standard Industrial Classification of Economic Activities from the Office for National Statistics data was used within this analysis. There were 19 sectors that were identified who utilise Scottish and Welsh income tax codes, and during the analysis and there were several schemes that couldn’t be matched, these are represented by blank or no sector found. Definitions for the sectors and their associated letters are shown in Table 2, and the split of sectors administrating Scottish/Welsh tax codes is shown in Figure 6.
Table 2: Definitions for sectors and their associated letters
| Description | Sector |
|---|---|
| Agriculture, Forestry and Fishing | A |
| Mining and Quarrying | B |
| Manufacturing | C |
| Electricity, Gas, Steam and Air Conditioning Supply | D |
| Water Supply; Sewerage, Waste Management and Remediation Activities | E |
| Construction | F |
| Wholesale And Retail Trade; Repair of Motor Vehicles and Motorcycles | G |
| Transportation and Storage | H |
| Accommodation and Food Service Activities | I |
| Information and Communication | J |
| Financial and Insurance Activities | K |
| Real Estate Activities | L |
| Professional, Scientific and Technical Activities | M |
| Administrative and Support Service Activities | N |
| Public Administration and Defence; Compulsory Social Security | O |
| Education | P |
| Human Health and Social Work Activities | Q |
| Arts, Entertainment and Recreation | R |
| Older SIC that was used as a placeholder for unspecified activities no longer standard practice | Z |
Figure 5: Percentage split of sectors administrating Scottish/Welsh tax codes for 2027 YTD (6 April 2026 to 6 August 2026)
| Sector | Sum of number of employments |
|---|---|
| Blank | 2,330,910 |
| No Sector Found | 8,45,271 |
| Financial and Insurance Activities | 731,022 |
| Professional, Scientific and Technical Activities | 571,334 |
| Administrative and Support Service Activities | 552,634 |
| Wholesale and Retail Trade; Repair of Motor Vehicles and Motorcycles | 460,316 |
| Real Estate Activities | 267,537 |
| Electricity, Gas, Steam and Air Conditioning Supply | 254,381 |
| Arts, Entertainment and Recreation | 231,639 |
| Transportation and Storage | 172,326 |
| Public Administration and Defence; Compulsory Social Security | 149,028 |
| Accommodation and Food Service Activities | 130,908 |
| Construction | 126,860 |
| Information and Communication | 104,294 |
| Water Supply; Sewerage, Waste Management and Remediation Activities | 37,803 |
| Agriculture, Forestry and Fishing | 26,184 |
| Education | 12,810 |
| Manufacturing | 11,783 |
| Mining and Quarrying | 3,259 |
| Older SIC that was used as a placeholder for unspecified activities no longer standard practice | 2,066 |
| Human Health and Social Work Activities | 204 |
Those not assigned to a sector accounted for the most discrepancies for both NPS Scottish or NPS Welsh, and the Financial and Insurance Activities sector, Sector K, accounts for the second most discrepancies for NPS Scottish or NPS Welsh, both shown in Figure 6.
Figure 6: Percentage split of sectors who show employment as NPS Welsh income code and RTI is not for 2027 YTD (6 April 2026 to 6 August 2026).
| Sector | Sum of Number of Employments |
|---|---|
| No Sector Found | 40,632 |
| Financial and Insurance Activities | 15,017 |
| Blank | 10,174 |
| Wholesale And Retail Trade; Repair of Motor Vehicles and Motorcycles | 4,685 |
| Transportation and Storage | 4,153 |
| Administrative and Support Service Activities | 3,973 |
| Electricity, Gas, Steam and Air Conditioning Supply | 3,693 |
| Public Administration and Defence; Compulsory Social Security | 3,133 |
| Professional, Scientific and Technical Activities | 3,048 |
| Construction | 2,096 |
| Real Estate Activities | 1,807 |
| Accommodation and Food Service Activities | 1,389 |
| Arts, Entertainment and Recreation | 947 |
| Agriculture, Forestry and Fishing | 564 |
| Information and Communication | 474 |
| Education | 441 |
| Water Supply; Sewerage, Waste Management and Remediation Activities | 83 |
| Manufacturing | 35 |
| Mining and Quarrying | 9 |
| Older SIC that was used as a placeholder for unspecified activities no longer standard practice 7 | |
| Human Health and Social Work Activities | 3 |
High-rate Taxpayers and Low-rate Taxpayers
For the purposes of this analysis, taxpayers were split into 2 bands, high taxpayer and lower taxpayer, the definitions are defined below in Table 3 and the split of taxpayer types assigned Scottish/Welsh tax codes is shown in Figure 9.
Table 3: High taxpayer and lower taxpayer bands
| Definition | Individual with a taxable income of |
|---|---|
| Scottish Higher Rate Taxpayer | £43,000 and over |
| Scottish Lower Rate Taxpayer | £42,999 and under |
| Welsh Higher Rate Taxpayer | £50,000 and over |
| Welsh Lower Rate Taxpayer | £49,999 and under |
| Scottish/Welsh Taxpayer | Individual not notably liable to pay tax |
For discrepancies for Scottish and Welsh income tax codes, Figure 10, low taxpayers account for most of the discrepancies, 87.38% and 89.12% respectively.
Figure 7: Percentage split of taxpayer types who have been assigned Scottish/Welsh tax codes for 2027 YTD (6 April 2026 to 6 August 2026)
| Tax Payer Category | Number of Employments |
|---|---|
| Scottish Low Tax Payer | 4,291,362 |
| Welsh Low Tax Payer | 2,302,812 |
| Scottish NNL Tax Payer | 110,283 |
| NPS Scottish Low Tax Payer | 89,483 |
| NPS Welsh Low Tax Payer | 85,878 |
| Welsh NNL Tax Payer | 62,636 |
| Scottish High Tax Payer | 33,390 |
| NPS Scottish NNL Tax Payer | 12,642 |
| NPS Welsh NNL Tax Payer | 10,355 |
| RTI Scottish Low Tax Payer | 8,941 |
| Welsh High Tax Payer | 7,442 |
| RTI Welsh Low Tax Payer | 4,758 |
| RTI Scottish NNL Tax Payer | 1,002 |
| RTI Welsh NNL Tax Payer | 767 |
| NPS Scottish High Tax Payer | 284 |
| NPS Welsh High Tax Payer | 130 |
| RTI Scottish High Tax Payer | 128 |
| NPS Welsh/RTI Scottish Low Tax Payer | 109 |
| NPS Scottish/RTI Welsh Scottish Low Tax Payer | 96 |
| NPS Scottish/RTI Welsh Scottish NNL Tax Payer | 26 |
| NPS Welsh/RTI Scottish NNL Tax Payer | 24 |
| RTI Welsh High Tax Payer | 14 |
| Neither Tax Code Present | 6 |
| NPS Welsh/RTI Scottish High Tax Payer | 1 |
Figure 8: Percentage split of taxpayer types who show NPS as Welsh income code and RTI is not for 2027 YTD (6 April 2026 to 6 August 2026)
| Tax Payer Category | Number of Employments |
|---|---|
| NPS Welsh Low Tax Payer | 85,878 |
| NPS Welsh NNL Tax Payer | 10,355 |
| NPS Welsh High Tax Payer | 130 |
No Net Liability
As mentioned, Operational Excellence began carrying out root-cause on the employers causing discrepancies. During this sampling period it was identified that No Net Liability, NNL, taxpayers, those individuals not notably liable to pay tax, could be leading to some of the discrepancies, so a flag identifying those individuals was added to understand what proportion of NNL employments there were, these are also shown in Figure 10.
Software
Software is defined as what Employers, Pension administrators and 3rd Parties use to file RTI data using either third-party software or, if they have 9 or less employees, they can use HMRCs free Basic PAYE Tools (BPT).
Additional Insight
During analysis, it was identified that the country code found on National Insurance and PAYE System, NPS, may be leading to some of the discrepancies. Further analysis has been commissioned to investigate whether this is one of the root causes for the discrepancies.
In addition, colleagues from Operational Excellence were given a random sample of 100 employers who had been issued a P6/P9 code and their associated employees causing the discrepancies. A control group of 100 different employers were chosen to allow to compare to begin root-causing. This information is not included in this report.
Country Code Analysis
Taxpayers are assigned a country of residence flag within NPS. Figure 10 shows the percentage split of employments’ assigned country code who have Scottish/Welsh tax codes assigned.
Figure 10: Percentage split of employments assigned country code who have Scottish/Welsh tax codes assigned for 2027 YTD (6 April 2026 to 6 August 2026)
| Country code | Sum of Number of Employments |
|---|---|
| Great Britain | 4,052,122 |
| Scotland | 1,773,526 |
| Wales | 911,314 |
| England | 192,191 |
| Not Specified or Not Used | 89,853 |
| Other | 3,157 |
| Northern Ireland | 381 |
| Isle of Man | 25 |
Figure 11 shows the percentage split of employments assigned country code who show NPS as Welsh income code and RTI is not. It shows that 66.51% of the employments who had NPS assigned a Welsh income code and RTI not had a Great Britain country code assigned.
Employments should not be assigned a Great Britain or United Kingdom country code, and analysis suggests these could be the reason for some of the discrepancies.
Figure 11: Percentage split of employments assigned country code who show NPS as Welsh income code and RTI is not for YTD (6 April 2026 to 6 August 2026)
| Country code | Sum of Number of Employments |
|---|---|
| Great Britain | 64,087 |
| Wales | 26,133 |
| England | 3,154 |
| Not Specified or Not Used | 2,854 |
| Other | 80 |
| Scotland | 44 |
| Northern Ireland | 11 |
Further Analysis
Some of the identified findings from the Operational contact with employers, filters for these have been added to the report but require further analysis is required into:
- which payments are causing the most discrepancies (such as one-off payments)
- whether more discrepancies are caused by employments or pensions
- whether more discrepancies are caused by after leaving payments
- whether No Net Liability taxpayers are causing the most discrepancies
A process improvement will be implemented in the coming months to prevent GB being used as a country code, to see if this will help solve some of the tax code discrepancies.
Definitions
Throughout the report the below definitions will be used:
| NPS | National Insurance and PAYE System (was previously called NIRS) is the strategic platform for income tax for individuals, primarily concerned with the administration of National Insurance and Income Tax via Pay As You Earn (PAYE), (the HMRC administered elements of) Student Loans, Claims cases, Child Benefit and Tax-Free Childcare eligibility. NPS is regarded as the master of citizen data for HMRC. |
| RTI | Real Time Information System is the HMRC system that employers submit payroll data through to HMRC |
| Scottish Income Tax Code | You pay Scottish Income Tax if you reside in Scotland for more than 183 days in a year (from 6th April of the given tax year). You may also pay Scottish Income Tax if you: move to or from Scotland, live in a home in Scotland and one elsewhere in the UK, for example for work, do not have a home and stay in Scotland regularly, for example you stay offshore or in hotels, rates show here. S1257L. |
| Welsh Income Tax Code | You pay Welsh Income Tax if you if you reside in Wales for more than 183 days in a year (from 6th April of the given tax year). You may also pay Welsh Income Tax if you: move to or from Wales, live in a home in Wales and one elsewhere in the UK, for example for work, do not have a home and stay in Wales regularly, for example you stay offshore or in hotels, rates show here. C1257L. |
| Scottish Tax Code Match | Both RTI and NPS Match as Scottish Tax Code. |
| RTI Scottish/NPS Welsh | RTI showing as Scottish Tax Code and NPS as Welsh Tax Code. |
| RTI Welsh/NPS Scottish | RTI showing as Welsh Tax Code and NPS as Scottish Tax Code. |
| Welsh Tax Code Match | Both RTI and NPS Match as Welsh Tax Code. |
| NPS Scottish/ RTI Not | NPS showing as Scottish Tax Code and RTI as regular Tax Code. |
| NPS Welsh/ RTI Not | NPS showing as Welsh Tax Code and RTI as regular Tax Code. |
| RTI Welsh/ NPS Not | RTI showing as Welsh Tax Code and NPS as regular Tax Code. |
| RTI Scottish/ NPS Not | RTI showing as Scottish Tax Code and NPS as regular Tax Code. |
| Other Tax Code | Tax codes that are not Scottish/Welsh. |
| Payment Level/Payments | The data used throughout the insight is at payment level, it is showing the tax code each time someone gets paid. This is split at their employment level, meaning if Joe Bloggs works at 2 companies and gets paid monthly, they will appear 24 times within a financial year worth of data. |
| Employment Level/Employments | If Joe Bloggs works at 2 companies, irrespective of their payment frequency as long as have received a payment for both jobs, they will appear twice within a financial year worth of data. |
| Occupational Pension (OPCN – Occupational Pension Contribution) | Used when any pension provider makes pension payments. |
| Scottish High Taxpayer | Earns £43,000 and over. |
| Scottish Low Taxpayer | £42,999 and under. |
| Welsh High Taxpayer | £50,000 and over. |
| Welsh Low Taxpayer | £49,999 and under. |
For this insight the size of the employers, pension administrators and 3rd parties were determined by how many staff they employ:
| Large | 250 employees or more. |
| Medium | 50 – 249 employees. |
| Small | 10 – 49 employees. |
| Micro | 1 – 9 employees. |
Annex C: Annual Business Intelligence Report 2025 to 2026
Customer Contact: Telephone
HMRC has a WRIT telephone route within the HMRC Personal Tax helpline. This gives customers generic, pre-recorded WRIT messages prior to speaking to an HMRC customer adviser. These figures do not represent all calls by Welsh taxpayers to HMRC.
Phone Calls Received and Answered
| Calls | 2025-26 | 2024-25 | 2023-24 | 2022-23 | 2021-22 | 2020-21 |
|---|---|---|---|---|---|---|
| Calls received | 6 | 21 | 35 | 87 | 92 | 120 |
| Calls answered | 5 | 17 | 23 | 63 | 70 | 83 |
Customer Contact: Complaints
HMRC tracks all WRIT customer complaints to ensure they are processed within the HMRC customer service targets. These figures do not represent all customer complaints received from Welsh taxpayers.
Customer Complaints and Response Targets
| Complaints and response target | 2025-26 | 2024-25 | 2023-24 | 2022-23 | 2021-22 | 2020-21 |
|---|---|---|---|---|---|---|
| Number of complaints | 0 | 21 | 35 | 87 | 92 | 120 |
| 15 day response target (80%) met | 0 | 17 | 23 | 63 | 70 | 83 |
Web pages monitored
HMRC has several WRIT-related webpages on GOV.UK and hits to these pages are monitored:
- Income Tax in Wales
- HMRC internal manual: Welsh taxpayer technical guidance
- Tell HMRC about a change in personal details, which is used by customers across the UK
Web hits
| GOV.UK web page | 2025-26 | 2024-25 | 2023-24 | 2022-23 | 2021-22 | 2020-21 |
|---|---|---|---|---|---|---|
| Income Tax in Wales | 17,338 | 32,314 | 38,953 | 44,444 | 39,827 | 43,337 |
| HMRC internal manual: Welsh taxpayer technical guidance | 42 | 67 | 75 | 186 | 178 | 283 |
| Tell HMRC about a change to your personal details | 596,495 | 754,848 | 1,072,096 | 1,307,931 | 1,332,067 | 1,530,708 |
Annex D: Metrics dictionary
This annex defines the headline metrics used in this report and explains their scope, source and any important caveats. It is intended to support consistent interpretation of the figures presented in Welsh Rate of Income Tax at a Glance and elsewhere in the report.
| Metric | Definition | Source | Limitations |
|---|---|---|---|
| Welsh taxpayers identified | The number of individuals recorded by HMRC as Welsh taxpayers for the relevant reporting period. | Derived from HMRC administrative records used to identify Welsh taxpayer status, including PAYE and Self Assessment related systems. | This is an administrative count rather than a definitive population measure of Welsh residents. Status depends on the address and residency information held by HMRC at the time of reporting. |
| Welsh Rates of Income Tax outturn attributable to the Welsh Government budget | The amount of Non-Saving/Non-Dividend Income Tax attributable to the Welsh Government Budget | Produced through HMRC’s outturn methodology using PAYE and Self Assessment data, with a small estimated component where final liabilities are not yet fully established. | A sm |
| WRIT Board meetings held | The number of formal Welsh Income Tax Board meetings held during the reporting year. | Taken from governance records for HMRC and Welsh Government. | This metric reflects meeting frequency only and does not, by itself, capture the scope or impact of the Board’s work. It also does not capture additional ad-hoc meetings. |
| Letters sent as part of the third-party data assurance exercise | The number of letters issued to individuals whose address information required confirmation or update following the third-party data assurance exercise. | Drawn from the annual third-party data assurance exercise and associated operational follow-up activity. | The number of letters sent is an activity measure. It does not, on its own, indicate how many records were subsequently corrected or how much tax was at risk. |
| Pieces of data analysis published | The number of analytical publications or formal analysis outputs on Welsh Income Tax published during the reporting year. | Based on HMRC analytical outputs published in support of WRIT governance, transparency or policy work. | This metric measures output volume, not the relative scale, complexity or policy significance of each publication. |
| Calls to the Welsh Income Tax telephone line | The number of calls to the WelshW Income Tax telephone line. | Taken from HMRC administrative data. | The figures do not represent all calls made by Welsh taxpayers to HMRC. |
| Welsh Self Assessment tax returns received | The number of Self Assessment tax returns received from taxpayers recorded as Welsh taxpayers. | Taken from HMRC Self Assessment administrative data for the relevant return cycle. | Return volumes may reflect timing effects, late filing and status changes. They do not represent the full Welsh taxpayer population because many taxpayers are dealt with solely through PAYE. |
| Guidance products updated | The number of guidance products reviewed and updated to reflect the WRIT rates, processes or supporting information applicable in the reporting year. | Drawn from HMRC content and guidance maintenance activity, including updates made ahead of the start of the tax year. | The count measures updated products rather than the scale of the changes made to each product. |
| Invalid or blank postcodes updated | The number of invalid or blank postcode records corrected through address assurance activity during the reporting year. | Taken from HMRC address assurance scan activity and related data quality corrections. | This is a data quality activity measure and should not be interpreted as the number of taxpayers incorrectly identified overall. |
| Number of postcodes updated | The number of postcode changes incorporated into HMRC’s residency and postcode reference processes, including new or revised postcodes received from the Office for National Statistics. | Based on quarterly Office for National Statistics postcode updates and HMRC process updates. | This metric reflects maintenance of postcode reference data rather than direct changes to individual taxpayer residency status. |
-
A Temporary Reference Number (TRN) is a number that is allocated by HMRC to records that do not have a matched or known National Insurance Number (NINO). TRNs take the format ‘NN LN NN NN’ (2 numbers, a letter, 5 numbers). TRNs are usually replaced when an individual applies for a NINO from DWP. However, a small number of individuals retain a TRN. ↩