Skip to main content

Since July 2019, we’ve published annual Land Transaction Tax (LTT) statistics. This includes data showing the numbers of land transactions that occurred in local areas across Wales, including data for:

  • local authorities
  • Senedd constituencies
  • National Parks
  • areas of deprivation
  • built-up areas

These statistics build on annual data previously produced by HMRC on Stamp Duty Land Tax (SDLT) for local authorities until 2018. HMRC administered SDLT in Wales until LTT replaced SDLT in Wales from 1 April 2018. The WRA collects and manages LTT for the Welsh Government.

We’ve seen increasing interest and use of our LTT data over the last 5 years, particularly second homes matters. The purpose of this explainer is to:

  • share further insights around some of the local area statistics we produce
  • describe how our statistics can and cannot be used

As a tax authority, and in keeping with the Code of Practice for Statistics, we can only collect data that enables us to fulfil our role of managing tax.

How we produce local area statistics

Alongside this explainer, we’ve published our data by local areas for April 2022 to March 2023. This provides detailed statistics on residential and non-residential transactions for local authorities, with statistics for only residential transactions for the other geographic aspects.

Breaking down counts of tax transactions and the associated revenue for each local area may appear relatively simple. But to do this, we must allocate transactions to each based on the location of property or land taxed. While this is possible, there are some complex scenarios. For example:

  • when there are multiple land items involved in a transaction, and they span several local areas
  • where the postcode of the land provided contradicts the local authority stated on the return or is not provided at all

The quality section for the release explains how we’ve resolved these issues, and this page discusses some of the challenges in interpreting these statistics.

The statistical article ‘Second homes: what does the data tell us?’, published by Welsh Government, also provides further advice on how to use LTT statistics.

Interpreting the statistics: factors to consider

The statistics provide numbers of transactions in a particular year. That means:

  • the data relate to property that has changed hands during that year and should not be mistaken for a count of all residential and non-residential property in any given area
  • with only 5 years’ worth of data, we cannot currently tell if these transactions are representative of the overall stock of property for each area. In short, any given year’s transactions may not be an unbiased sample of that overall stock

However, we can use the data to highlight the extent of variation in the value of properties and the tax raised per transaction in each Welsh local authority area.

We can also analyse which types of transactions take place in each area. For example, there are different patterns for residential and non-residential transactions across individual local authorities in Wales.

By relating residential transactions with sources that estimate the stock of residential property, such as the number of council tax dwellings (local authority data) or council tax stock of properties data (Valuation Office Agency), we can also identify areas where, for example, there are comparatively more transactions than the average.

There’s been increasing interest in understanding the different LTT rates and bands and how and when higher residential rates apply. The next section outlines the guidance around higher residential rates and explains when and how they apply.

Higher residential rates of LTT: charges and when it applies

There are multiple reasons why the higher rates of tax apply but the 3 main reasons are:

1. Companies

Higher rates of LTT apply to any company or organisation purchasing residential property for any purpose such as for letting purposes, or for redevelopment. However, if certain conditions are met, a tax relief applies to some company cases that may fully or partially cover their liability when purchasing residential property.

In this context, examples of relieved transactions include:

  • movement of property between different parts of the same group of companies
  • purchases by social housing providers
  • companies providing certain alternative finance mortgages
  • part exchange purchases by housebuilders and in other circumstances

These transactions still form part of the higher rates transaction counts.

2. Not a main residence

Higher rates of LTT apply to all individuals who buy property that’s not to be their main residence, including:

  • residential property that’s not used by the owner as their main residence. For example, a residence used for weekends, for holidays, on a seasonal basis or during the working week to facilitate access to a place of work
  • homes bought to rent to other people for short and long-term use (including as residential or holiday lettings). This could be combined with occasional use by the owner as well
  • homes acquired for relatives, such as elderly dependents, children or students
  • homes bought to be developed or renovated for onward resale

3. Outstanding bridging cases

Higher rates of LTT apply to an individual buying a property as a new main residence, where they’ve not sold their previous residence at the time of purchase. But if the taxpayer sells their former main residence within 3 years, the individual is eligible for a refund of the higher rates element of the tax. In these cases, the transaction is changed to main rates residential at the point the refund is claimed.

In any given year’s statistics, only some refunds will have already been claimed. So, some higher rate transactions will eventually leave the count in subsequent years. We will not know the full extent of the outstanding bridging in the latest year for up to 4 years, including the allowed additional year for the claim itself.

We publish a range of statistics for higher rates transactions, including the total revenue raised and the number of refunds made each month. At a local area level, we publish the annual percentage of residential transactions for which the higher rates apply.

Using higher rates of LTT data accurately

As explained above there are different reasons as to why higher rates of LTT are applied. These data are sometimes misinterpreted, when used to comment on and highlight differences in second home ownership across Wales. Why may this not be appropriate?

Firstly, there’s not a single accepted definition of a second home, and different definitions are often relevant in different circumstances. We can broadly consider 2 scenarios:

  • a wider definition applies and may include any property not lived in by the owner as their main residence. But it may be lived in by others, such as private renters, or other family members
  • a narrower definition of properties used only by the owner for only part of the year

There’s also the added complication of how using the property as part of a holiday letting business might interact with these definitions.

Secondly, the system used to administer the tax does not need to directly distinguish between the reasons for charging higher rates. It simply asks whether the higher rates apply. It’s not currently possible to directly identify these elements within the data, although we can do something here, as explained later.

Thirdly, the fact that these statistics only apply to changes each year - and may not be representative of all properties in the area - is pertinent, for 2 reasons:

  • only a relatively small share of residential properties (less than 5%) changes hands each year, and with only 5 years’ worth of data so far, it is unknown whether higher rates properties transact more often than other residential transactions. If that’s the case, it’s unlikely that conclusions drawn from that higher rates data can be reliably applied to the entire stock of property in each area
  • we cannot routinely tell whether properties were already in one of the categories to which higher rates apply before the transaction. As such, it’s not possible to say that transactions in a year changed the ownership profile of the property stock of an area. For example, a transaction may already be a holiday home sold to a person who will again use it as a holiday home, or sold from one buy-to-let landlord to another

And finally, when considering breakdowns of higher rates transactions, some will inevitably be averaged out within different geographical areas. For example, when looking at local authority data, there will be areas where the categories described above apply to differing degrees, and the effect of any single factor may be reduced across the entirety of the local authority. However, it’s worth noting that some parts of Wales are likely to have relatively high concentrations of more than one factor. For example, those with both student populations and popular tourist locations.

Using the data accurately: advice

Any use of these statistics should reflect the fact we’re referring to ‘in-year’ transactions of properties and not the total stock of properties, as well as being clear about what’s included. Ideally, the data should be combined with other information. For example, knowledge of the private rented sector market in an area. Referring to both data sets will create a clearer picture than using our LTT data in isolation.

If the LTT data are being used in isolation, however, we’d advise that this be done carefully considering the information we publish. We know already that most of the higher rate transactions fall into the ‘bought for use other than as a main residence by the owner’ category, whether bought by an individual or a company.

How to express the statistics: examples

We can legitimately use the data to make statements like:

“In 2022 to 2023, … properties in … were subject to the higher rates of LTT, most of which were bought for use other than as a main residence by the owner.”

“In 2022 to 2023, … was the area with the highest proportion of sales subject to the higher rates of LTT. Most of those were bought for use other than as a main residence by the owner.”

However, it’s not accurate to use the data to make statements like:

“…% of property in … are second homes”.

This is incorrect on 2 counts because:

  • it makes an assertion about all property, not just those which were sold in-year
  • it incorrectly states that all higher rates transactions relate to second homes

This is especially problematic where the term ‘second homes’ has different meanings in different contexts.

It’s best not to make assertions around higher rates transactions changing the profile of stock in an area. This can only be done reliably with estimates of the whole stock, because as stated earlier, it’s quite possible that many of the higher rate transactions were in categories to which higher rates applied anyway before the transaction took place. 

For users particularly interested in the stock of residential property, local authority council tax dwellings data is a useful source of information at local authority level.

Can we improve our data?

We recognise that there’s appetite for these data. As such, we’ve started to experiment with the data.

1. Analysing the effect of outstanding bridging

A person can reclaim the higher rates of tax if they dispose of their former main residence within 3 years of the date they acquire their new main residence, but most will occur within the first or second years. Some of those refunds will have already been made during the latest financial year, and so we use the term ‘outstanding bridging’ to represent those refunds still to occur in subsequent years. Therefore, these are the transactions we’d expect to eventually leave the counts of higher rates transactions, when they change to main rate residential transactions at that point.

In previous years, we’ve used data on refunds for transactions effective in 2018 to 2019 to model the effect of outstanding bridging for the latest financial year. Within those original estimates, there’s an assumption that the rates of refund stay constant across years. This assumption is not fully supported upon investigation, with further years of data now available.

Therefore, starting in the previous version of this article, we adapted the methodology for outstanding bridging to use existing data on all refunds (excluding refunds approved during the same financial year that the original transaction was effective). We’re more confident both with the accuracy of these latest estimates, and of the relevance in this context of estimates excluding those already refunded.

2. Breaking out company data

We’ve split the company purchase element out from the data and reported this using 2 categories:

Purchases relating to the public sector and certain relieved transactions

These mainly relate to social housing but also include

  • those which relate to intermediate temporary purchase, such as certain alternative mortgage providers or part exchange by housebuilders
  • those which are moved around companies, all of which are generally relieved

Everything else (in terms of company purchases)

This is more likely to be relevant to use in the buy to let sector but may also be relevant in the second or holiday home context. However, we only have limited information currently for this data.

3. Individuals buying property for use other than own main residence

By excluding from the total higher rates transactions:

  • an estimate of transactions which are expected to be bridging
  • the 2 types of company purchases, as estimated above

we’re essentially left with individuals buying property for uses other than their own main residence. We can now refine the first of the statements made earlier to be more direct, for example:

‘In 2022 to 2023, approximately … properties in … were bought by individuals not using it as their own main residence’ 

Note the use of the word ‘approximately’. We’d always recommend using this as there’s a level of estimation involved.

The chart below summarises how this extra detail is reflected in the 2022 to 2023 data we published in our latest annual LTT statistics.

The attached spreadsheet sets out all the data and provides fuller details around the methods used.

Chart: Approximate percentage of residential transactions in different LTT higher rates categories, 2022-23

Details are in the text following the chart.

 The bar chart shows the different types of higher rate transactions stacked as a percentage of all residential transactions for each local authority and Wales. Gwynedd and Isle of Anglesey have the highest percentage of individual purchases not bought as main residence, and Blaenau Gwent and Merthyr Tydfil have the highest percentage of company purchases.

Source: Data used in Welsh Revenue Authority local area statistics 2022 to 2023 explained (55 Kb, Open Document Spreadsheet)

What’s next?

We’ll continue to update this explainer annually. In time, one area we’d like to investigate is further separation within:

  • the ‘bought by individuals for uses other than as a main residence’
  • the ‘wider company purchase’ elements of the data

In June 2023, we introduced a new question on the LTT return. This question asks about the intent behind residential purchases at the higher rates of tax, including whether they are second home purchases or buy-to-let properties. In conjunction to this, we currently match our property data to existing data sources such as data held by Rent Smart Wales for operational reasons.

Using these data sets may enable us to provide a clearer picture of the number of second homes and buy-to-lets each year. We will review the quality and usefulness of this data going forward, and provide an update in the summer 2024, alongside the updated annual LTT statistics.

We’d also like to extend the modelling to our other geographic datasets or possibly different groupings of local areas based on other common characteristics.

Your feedback

If you think any of this further analysis will be useful or, have any comments on the data we already publish on either LTT or Landfill Disposals Tax, we’d be interested in hearing from you. 

If you have any comments or queries about using our data, please email us