Official Statistics

Quality report: Annual Tax on Enveloped Dwellings (ATED) statistics

Updated 10 May 2023

1. Contact

Organisation unit: Knowledge, Analysis, and Intelligence (KAI)
Name: Indirect Tax Stamps & Property, Indirect Taxes, Customs & Co-ordination
Mail address: HM Revenue and Customs (HMRC), 100 Parliament street, London SW1A 2BQ
Email: stamptaxes.statistics@hmrc.gov.uk

2. Background to the statistics

This methodology and quality report relates to the Annual Tax on Enveloped Dwellings (ATED) statistics publication. This publication relates to the total amount of receipts, number of liable declarations and types of reliefs claimed.

The purpose of this document is to provide users of statistics with information about the quality of the outputs, measured against difference dimensions of statistical quality, as set out by the Code of Practice for Official Statistics.

3. Coverage

3.1 Time period and geographic regions

This publication series provides an annual view of the receipts, number of liable declarations made and types of reliefs claimed for the Annual Tax on Enveloped Dwellings (ATED) from its introduction in the 2013 to 2014 financial year. This is a UK wide publication as this type of taxation has yet to be devolved to separate parts of the United Kingdom.

3.2 Data splits provided in the publication

Data is broken down by ATED receipts, declarations made, geographical region, local authority and type of relief being claimed since the 2013 to 2014 financial year.

3.3 Future coverage information

The geographical and local authority data being reported is based entirely on the reported postcode. Where this was entered incorrectly or had been omitted then it was not possible to identify the specific address area. This does not impact on the values being reported for receipts, declarations or reliefs being claimed.

4. Data sources

4.1 Returns and declaration data

The information in this publication is based on HMRC’s records of submitted ATED liable returns and Relief Declaration returns. Due to the transition from paper returns to online returns, receipts data for the 2016 to 2017 financial year, and reliefs’ data from the 2015 to 2016 financial year onwards, are taken from two different data sources for each type of return. This was no longer the case for the 2021 to 2022 financial year where there was only one source for this. Declarations are initially identified by the year they relate to (not when the declaration was made). Where a person submits multiple returns, data from the most recent return are used. We also check and filter for duplicate returns between filing systems. The aggregate revenue figures (including breakdowns) are then constrained to the financial accounts published in HMRC’s annual report, meaning receipts are counted on the basis of when they are received, not when they are due. In line with HMRC’s statement on statistical quality published in ‘HMRC: Official Statistics’.

These statistics are based on an administrative data source collected by HMRC in the running of its business (assessment and collection of tax). Any apparent discrepancy between totals and the sum of the constituents in the publication is due to rounding.

The quality of these statistics depends on the purpose that they are intended to be used for. The definitions used within this publication are defined by tax law, and whether or not a property is counted within the estimates is dependent on the submission of an ATED Return. For example, properties where no return is required, or is required but not submitted, will not be counted in the provisional figures.

Please see Annual Tax on Enveloped Dwellings: Returns Notice for more information.

4.2 Relief’s data

From 1 April 2015 a simplified return (the Relief Declaration Return) was introduced for properties entitled to claim full relief from ATED and for which there is no ATED to pay. On these returns property value information is not collected, meaning reliefs cannot be split into bands as in previous publications.

5. Receipts and transactions methodology

Data is extracted from HMRC’s live database which stores the returns data and is stored in the Corporate Data Warehouse mainframe. New returns are added to the existing database, and records from returns which have been updated (due to voided transactions, errors, refunds or customer-generated updates) are amended. Automated data validation is performed at this point, removing duplicate records.

Duplicates can occur when a transaction has accidentally been submitted twice (online). Instances where there are multiple entries per transaction (such as details for more than one purchaser) are filtered to make a single record per transaction.

HMRC statisticians access this extracted data using SAS, SAS is a statistical analytics software developed by the SAS institute. Snapshots of the ‘live’ database are taken on a monthly basis for producing internal management information, and at this point statisticians perform basic data validation which include the following:

• checking for outliers

• checking record counts and total tax against previous months and expected values including fiscal forecasts

Any records identified as suspicious will either be amended where the error is obvious (for example decimal point errors) or removed where they present a risk to the accuracy of the published statistics. Records are also removed where their inclusion presents a risk of taxpayer disclosure.

The final snapshot of the live database is taken several months after the end of the reporting year. This dataset is then cleansed and modified (mainly to identify specific geographic areas from the reported postcode) to remove questionable values and in certain instances, then advice from policy/operational experts within HMRC is sought to remedy potentially erroneous entries.

Standard queries are then run to produce aggregated data, which is then arranged into output tables for publication. These tables undergo several rounds of quality assurance and scrutiny, including a Quality Assurance (QA) checklist, to ensure the outputs are accurate and consistent, before being published on an annual basis.

6. Presentation of data and revisions

6.1 Rounding:

Estimates of ATED receipts are rounded to the nearest million pounds. Whilst declaration numbers for regions, local authority, relief claims are rounded to appropriate levels. This is done both for presentational and also dominance and disclosure policy purposes. Some rounded figures may not sum to the full amount stated elsewhere due this rounding convention.

6.2 Seasonal Adjustments

As this is an annual publication there are no seasonal effects at play.

6.3 Revisions and Timing of Publication

This annual publication is produced on a financial year basis and is updated every March/April with data from the previous financial year. This provides enough time for the data to be processed and cleaned, and for the publication to be constructed. To note, a clerical error was found with the calculation of the 2016 to 2017 financial year declarations (but not the receipts figure) estimate. This was been shown as a revision in the values reported in subsequent reports.

6.4 Geographic Breakdowns:

The split of revenue between the various geographic units is based upon the location of the property, such as the postcode recorded on the ATED return. Due to difficulties with the introduction of the new data source there are a small minority of cases where this data is not available. This is an issue that is currently being investigated. Additionally there are a further number of cases where data is not fully recorded. Where these cannot be corrected, they are classed as ‘missing’ and assigned to the ‘missing’ category in the tables. The aggregation of the ATED revenue in each table is constrained to the total revenue collected. Office for National Statistics (ONS) lookup tables are used to match postcodes to appropriate UK geographic areas. More information can be found on the GOV.UK website regarding presentation guidance for administrative areas.

6.5 Disclosure Control

HMRC reviewed its procedures around disclosure and taxpayer confidentiality in 2013. As a result, we are narrowing the scope of the estimates we publish in this release to ensure that we are fully complying with the National Statistics Code of Practice and our obligations under Section 18 of the Commissioners for Revenue and Customs Act 2005 (CRCA), which makes it clear HMRC must not disclose taxpayer information to anyone, unless there is lawful authority to do so. Therefore, some local authority data is combined with a near neighbour or not released at all. We will keep this issue under regular review. More information can be found on the GOV.UK website for Annual Tax on Enveloped Dwellings returns guidance.

6.6 Publication Calendar

Publications will be released in either March or April each year, but may be adjusted whenever the day preceding the last working day is a non-working day (publications are never published on a Monday, Saturday or Sunday).

The full publication calendar can be found on the ‘scheduled updates’ page of Statistics at HMRC on GOV.UK. It is also published on the Publication Hub for UK National Statistics.

The statement of compliance with the Pre-release Access to Official Statistics order and the Pre-release access list can also be found on the HMRC website:

• HMRC Pre-release statement

• Pre-release access list

6.7 Previous versions

As ATED was only introduced in the 2013 to 2014 financial year, this is only the second version of the quality report that has been published.

7. Assessment of quality

7.1 Relevance

The statistics primarily provide information on ATED transactions on an annual basis from the 2013 to 2014 financial year. The number of transactions, ATED receipts and relief claims are broken down by local authority and geographical region.

These outputs can be used to examine the effectiveness of this taxation in countering potential avoidance of the payment of Stamp Duty Land Tax.

The publication is primarily used by analysts, commentators and other government departments to examine UK trends in the housing market. They are also used to examine the impact of policy decisions and economic conditions on tax receipts.

7.2 Accuracy and reliability

The statistics are based on administrative data used by HMRC to run its business (collecting tax). They are therefore accurate and timely. The main quality issues are around completeness of information as much of the data is based on returns.

There are likely to be some inaccuracies in the data due to reporting or keying errors, such as misclassification or missing cases. As some of the data matching between different datasets require manual interpretations of available data.

7.3 Timeliness and punctuality

The release timetable is designed to provide users with the best balance of timeliness and accuracy. Release dates are announced in advance on the HMRC website.

This is an annual publication. The statistics show that data from the previous year are generally published on or near the last working day of March after the end of the year being reported on (unless that day is a Monday). The delay in publication is to allow time for most ATED returns to be submitted and for data cleansing.

7.4 Accessibility and clarity

The publication is available in HTML format and includes contact details (both email and telephone) for sending queries and feedback to the production team.

The statistics are taken directly from the source data that is collected for administrative purposes with little manipulation between source and publication. Our publications directly reflect the tax system definitions and therefore the administrative data are an accurate and relevant data source.

7.5 Coherence and comparability

Departmental performance, such as the annual report and the Trust Statement. Total receipts suggested by the ATED returns are adjusted to match HMRC receipts as reported in the monthly HMRC Receipts and National Insurance contributions for the UK. There is often a few percentage points difference in the two sources, which is largely caused by the timing of payments (where purchases are completed late in the month, they may pay HMRC in the following month).

7.6 Trade-off between output quality components

The statistics are produced and released according to a preannounced timetable.

The timetable for publication is designed to provide users with the best balance of timeliness and accuracy. At the time of production of the publication, some records may still be incomplete, and compliance checks by HMRC may result in the subsequent alteration of returns.

7.7 Assessment of user needs and perceptions

The producers work closely with key customers and stakeholders in HMRC, HM Treasury and the Office of Budget Responsibility to keep track of changes and developments in the tax policy, and continuously review the coverage and content of our publications to ensure that they meet the needs of those users.

Broader consultations are conducted when appropriate, for example when significantly changing the provision or coverage of the published data, or revising the methodology used. These will generally involve contacting known users of the published statistics to ask specific questions or request feedback. These questions would also be published on the website adjacent/linked to the publication in order to capture users with whom we have had no previous contact.

The results of user consultations would then be published on the website. The latest consultation took place in 2015 and the results can be found alongside the most recent publications.

Contact details are available at the bottom of each publication for users to get in touch if they have a query. We also added some footnotes to explain changes in data and trends, as well as key definitions throughout the commentary.

7.8 Confidentiality, transparency and security

The data is held in a secure server. Access to this data is only granted to those who require it for business purposes only.

All persons that have pre-release access to the published data (aside from the production team) are listed on the HMRC website.

7.9 Performance, cost and respondent burden

All published tables use HMRC administrative data so there are no explicit costs of production. Much of the data would be produced regardless of publication for internal monitoring and policy development purposes.