Data used in Welsh Revenue Authority performance reporting 2022 to 2023

The Welsh Revenue Authority (WRA) published its 2022-23 annual report on 14 September 2023, including a section covering the performance of the organisation over the year (the “performance report”). The intention of this report is to provide a statistical view of the data used in that document, and to make the underlying datasets fully accessible for further reference or analysis.

There are several charts used in this report, presenting new data for each of the different performance indicators for the WRA for the period 2022-23. Some of these are included somewhere in the annual report, and referenced accordingly, although not necessarily in the same order. This report also contains some data not presented in the annual report to provide a more complete picture of all the data used by the WRA in its internal performance analysis.

For the purposes of viewing the data, it is sometimes easier to visualise reverse values of the performance indicators. For example, the percentage of returns filed on time is close to 100% and any variation when seen on a scale of 0-100% is very hard to see in a static image. Rather than changing the scale to say 97-100%, which will over-emphasise the variation, we instead present the percentage of returns which are NOT filed on time, using a scale of 0-3%. This gives a fairer reflection of the trend in a static image, and where this technique is used, the data for both the reverse measure and the actual measure are available in the accompanying spreadsheet.

This report includes brief analysis of each measure, and the reader may wish to consider this in conjunction with the more detailed narrative in the performance report, which is set in the context of the organisation’s approach and objectives.  Where applicable, bookmarks to the relevant parts of the annual report are added under each chart.

Note that most of the analysis below applies to Land Transaction Tax (LTT), although where it is also relevant to include Landfill Disposals Tax (LDT) data, that data are also included in the measure.

Chart 1: Number of LTT returns submitted to the WRA, by month received

Image

This chart is not shown explicitly in the related Annual Report.

This chart sets out the context against which the performance measures should be considered. It shows how the numbers of LTT transactions received in each month has changed over the course of April 2021 to March 2023.

The chart shows that throughout the 2022-23 financial year transactions were slightly lower than in 2021-22.  Transaction numbers peaked in June 2021, as the temporary reduction in Land Transaction Tax first introduced in July 2020 came to an end, with some transactions brought forward to take advantage of the reduced tax.  Numbers then reduced to around 6,000 per month over the latter part of 2021, before settling at around 5,500 over the course of 2022. Seasonal dips were seen in January and February of both 2021 and 2022.

Chart 2: The percentage of transactions that are not processed automatically through to initial payment, by month received

This chart is not shown explicitly in the related Annual Report.

The WRA has a performance target of 98% of transactions automatically processed with no manual involvement.  This covers both the receipt of digital transactions and automatic matching of payments received to any transactions where there is a financial liability and covers both LTT and LDT.

This chart uses the reverse technique explained in the introduction and shows that the percentage of transactions not processed automatically generally decreased over the 2022-23 financial year, continuing the trend that was also evident in the previous year. This measure still has some way to go to reach a relatively ambitious target, likely to require improvements in the automatic matching of payments currently being considered by the WRA.

Chart 3: The percentage of Land Transaction Tax returns received outside 30 days, by month transaction was effective

This chart is not shown explicitly in the related Annual Report.

The WRA has a performance target of 98% in respect of receiving LTT returns on time, that is within 30 days of the effective date.
This chart uses the reverse technique explained in the introduction and shows that the percentage of returns received after 30 days of the effective date has been within the target range since January 2022, with a general downward trend over the last two years. This measure generally improved in the last 6 months of 2022-23, coinciding with a decrease in the overall number of transactions.

Chart 4: The percentage of Land Transaction Tax returns paid outside of 30 days, by month transaction was effective

This chart is not shown explicitly in the related Annual Report.

The WRA has a performance target of 98% in respect of receiving payments against LTT returns on time, that is within 30 days of the effective date.

This chart uses the reverse technique explained in the introduction and shows that the percentage of returns paid after 30 days of the effective date has generally decreased over the past 2 years, to touch our target line in February 2023. This fall has coincided with organisation‑wide and continuing efforts to support taxpayers in preventing and managing debt.

Chart 5: The percentage of Land Transaction Tax debts collected outside 30 days and 90 days, by month transaction was effective

This chart is not shown explicitly in the related Annual Report.

For each LTT transaction with a financial liability that is not both submitted and paid within 30 days, a debt is created. The WRA aims to collect payments on transactions that become a debt within 30 further days and has a target of collecting 90% of those debts in that timeframe, as well as a target of collecting 98% of those debts within a 90 day timeframe. Transactions that are submitted outside these timeframes create a bias in this measure and are excluded from the calculation. Instead, the WRA analyses these cases separately to ensure comparable timeframes are achieved following receipt.

This chart uses the reverse technique explained in the introduction and shows that debts collected within 30 days were within or near the target range for the majority of 2021-22 and 2022-23, with some peaks in July 2021 and November 2022 influenced by temporary stops on collection and delivery of post. The November 2022 spike coincided with a halt in managing post over Christmas, with performance quickly returning to more normal levels a month later.

For both these measures, particularly debts collected outside 90 days, the numbers of transactions can be low and are naturally highly variable. This is something that has been exacerbated by the drop in overall levels of debts seen in the previous chart, and it is important to consider the longer-term trends, both of which are broadly flat.

Chart 6a: The number of and mean average days to pay Land Transaction Tax higher rate refunds, by month of request

This chart is not shown explicitly in the related Annual Report.

Chart 6b: The percentage of Land Transaction Tax higher rate refunds paid outside 30 days and 60 days, by month of request

This chart is not shown explicitly in the related Annual Report.

Generally, when a taxpayer purchases a new main residence without selling their former main residence at the same time, they are subject to the higher rate of residential LTT. Those that sell their former main residence within three years are usually eligible for a refund of the difference between the higher rates and main rates LTT on the original transaction.

The WRA aims to make payments of these refunds, termed higher rate refunds, in as timely a manner as possible, and in any case within 30 days of receipt of the request. We now have three years’ worth of data on this basis.

The two charts above show how the time taken to repay higher rates refunds has changed over time, and the proportion of all refunds that are paid within both 30 and 60 days.

Chart 6a shows that the average time taken to process higher rates refunds decreased from a high in early 2021-22, before stabilising between 6 and 14 days. This stabilisation coincided with improvements to processes and re-organisation of staff resource. The proportion of refunds handled within 30 days and 60 days is shown in chart 6b, which uses the reverse technique explained in the introduction. As might be expected the low average time to process refunds leads to lower variation in this measure over the past 2 years, with only isolated peaks. These peaks are generally influenced by small numbers of cases where the delay is often due to additional information being sought, and slow taxpayer response.

Chart 7a: Average days to pay Land Transaction Tax general refunds, by month of request

This chart is not shown explicitly in the related Annual Report, but the data are referred to in the section titled Timeliness of WRA repayments.

Chart 7b: The percentage of Land Transaction Tax general refunds paid outside 30 days and 60 days, by month of request

This chart is not shown explicitly in the related Annual Report, but the data are referred to in the section titled Timeliness of WRA repayments.

The WRA’s 2022-2025 Corporate Plan contained a commitment to expand our reporting on refunds to include general repayments, alongside higher rate refunds. These repayments might pertain to duplicate payments, subsequent claims for relief, or overpayment following changes in tax bands. Although higher rate refunds constitute the majority of our refunds, these more general repayments account for the remaining 20%.

The annual report explains why we aren’t applying a target to these general repayments, mainly due to challenges with matching repayment requests to the repayments themselves. However, we’re monitoring the data and we present it here for transparency.

The two charts above show how the time taken to make these general repayments has changed over time (chart 7a), and also and the proportion of all refunds that are paid within 30 days and within 60 days (chart 7b, which uses the reverse technique explained in the introduction).

Following a high in early 2021-22, when we were dealing with a historically high number of higher rate refunds, the percentage of general refunds repaid outside 30 days has remained below 20%, with some fluctuation. Improved performance in the 6 months since October 2022 coincides with the WRA processing repayments in relation to a change in tax bands, which are typically more straightforward cases.

Chart 8: The percentage of transactions not paid correctly first time, by month transaction was effective

This chart is shown as chart 1 in the related Annual Report, and the data are referred to in the section titled Calculation services – right first time.

The WRA aims to capture more information about how easy it is to use our services, and our Corporate Plan 2022-2025 includes a new measure aimed at doing that. This is the percentage of transactions paid correctly first time, which means those paid in full in a single payment. We exclude transactions where an amendment to the tax due has been made as it impacts the number and level of payments for reasons not related to ease of use of our services.

Chart 8, which uses the reverse technique explained in the introduction shows how the cases where transactions were not paid correctly first time has generally decreased, indicating positive progress in this area.

Chart 9: Number of transactions and total tax in LTT risks 1 to 5

This chart is shown as chart 2 in the related Annual Report, and the data are referred to in the section titled Our approach to managing tax risk.

The WRA is able to carry out detailed analysis on the data contained in each transaction it receives to check for the presence of different characteristics that may indicate common errors or risks in the information provided. Each of those ‘tax risks’ are then analysed separately to identify the numbers of transactions and tax at risk, so that this can be tracked over time. The performance chapter of the annual report explains more about the WRA’s approach to managing tax risks.

Attention was focussed on several tax risk areas and we have reported on the 5 main tax risks here The individual datasets for each of those are shown in charts 10-14 below, whilst this chart shows the aggregation of the data for those 5 tax risk areas up to and including data for 2022-23. Data are available for 4 years for both the count of transactions (shown by the line, left hand axis) and tax at risk (shown by the bars, right hand axis).

These measures are further split into transactions which are probable risks, where work has been done to understand the potential risk, and unclear risks, where there still isn’t enough information to decide whether the transaction is risky or not. A consistent effort was made in the last 6 months of 2022-23 to reduce the quantity of unclear risks, as can be seen in the chart.

Following a general increase in the number of transactions falling into these 5 tax risks as shown by the line (left hand axis) during 2020-21, the number of transactions levelled off for most of 2021-22, but with a fresh increase in the last quarter of 2021-22. These numbers then slowly began to fall in 2022-23, before sharply decreasing in the final quarter of 2022-23. The related value of the tax at risk in those transactions shown by the bars (right hand axis) changed more rapidly than the number of transactions, due in the main to recent rises in the value of transactions. More explanation of this data in the context of the WRA’s approach and objectives is given in the annual report.

Chart 10: LTT risk 1 - companies buying residential property

Chart 10a: LTT risk 1a - residual activity around companies buying residential property

Please see the commentary under chart 9.

Following a sustained fall in 2019-20, the general risk was mitigated. However, some residual activity started in 2020-21, which can be attributed to a specific type of transaction is now shown separately in chart 10a. This risk can be seen to be generally minimal over the course of the past few years, with a steady number of transactions in 2022-23, similar to that of the previous year, but generally less than some of the other risks presented here.

Chart 11: LTT risk 2 - outstanding tax return

Please see the commentary under chart 9.

There was a general fall in transactions during the first two years from an early peak, into the first quarter of 2020-21. Since then, the risk grew until the first quarter of 2021-22, before decreasing and then stabilising over the next 2 years. The value of tax at risk has been affected by a few larger unrepresentative transactions over the period shown but largely tracks the numbers of transactions for this risk. The level of risk is generally less than some of the other risks presented here.

Chart 12: LTT risk 3 - disagreeing with LTT calculator

Please see the commentary under chart 9.

Since a low in the first quarter of 2020-21, cases increased in line with increasing numbers of transactions. During 2021-22, cases remained broadly stable apart from a small dip in quarter 3, while the value of the tax at risk went up a little at the start of 2021-22, before dropping back. In 2022-23, the number of transactions and the tax due on those transactions both decreased in the first 2 quarters, before remaining stable and low over the next 2 quarters. The level of risk is generally less than some of the other risks presented here.

Chart 13: LTT risk 4 - tax treatment of different property types

This chart is shown as chart 3 in the related Annual Report, and the data are referred to in the section titled Our approach to managing tax risk.

Please see the commentary under chart 9.

This is the largest of the risks covered in both terms of numbers of transactions and the related value of tax at risk associated with those transactions.

Since the low in early 2020-21, there was a marked increase in numbers in the second half of both 2020-21 and 2021-22, in excess of the numbers that might be expected as overall transactions increased, which suggests an increase in the risk itself. The value of tax at risk followed a similar pattern, although the data are also impacted by rising transaction values, with the figure in quarter 4 2021-22 being the highest seen to date by some margin.

After this peak, the number of transactions and tax due in this risk stabilised for the first 3 quarters of 2022-23, before dipping considerably in the 4^th quarter. This coincided with a sharp drop in the overall number of transactions, but even after allowing for that the level of risk has decreased. This drop in risk has itself coincided with changes made to our tax system to request further information about the transactions and additional efforts to educate taxpayer agents.

Chart 14: LTT risk 5 - in relation to a specific relief (relief a)

This chart is shown as chart 4 in the related Annual Report, and the data are referred to in the section titled Our approach to managing tax risk.

Please see the commentary under chart 9.

The count of transactions at risk rose since the beginning of 2021-22, broadly in line with the rise in general transactions, although the rise in the latter half of 2021-22 was steeper than expected, suggesting an increase in the risk itself. The count then reduced in the first quarter of 2022-23, stabilised for a while, before reducing again in the final quarter of 2022-23. As with the previous risk, this was due in part to a decrease in overall transactions, but also to a fall in the risk itself, coinciding with additional education efforts with taxpayer agents and the collection of further information about the transactions.

The value of tax at risk follows a similar pattern, although the data are also impacted by rising transaction values, with the figures in quarters 3 and 4 of 2021-22 and quarter 3 of 2022-23 being the highest seen to date by some margin.

Chart 15: Number and value of cases where LTT was protected

This chart is shown as chart 6 in the related Annual Report, and the data are referred to in the section titled Our approach to managing tax risk.

Over the 2022-23 financial year the WRA invested more effort into protecting tax. This is where taxpayers have submitted a downwards amendment to the LTT due on a tax return, but where we believe they were incorrect to do so. In such cases, we may open an enquiry and are usually successful in protecting tax that would otherwise have been refunded inappropriately.

The tax value of these cases are often highly variable and impacted by large individual transactions. However, chart 15 shows a significant increase in activity in this space in 2022-23, with virtually all cases resulting in protection of the LTT originally deemed due.

Chart 16: Number and value of cases where LTT was recovered or returned

This chart is shown as chart 5 in the related Annual Report, and the data are referred to in the section titled Our approach to managing tax risk.

For cases that fall into the risks identified in the charts above, the WRA will often open an enquiry into the LTT return, which may result in an amendment, usually upwards, to the return. In these cases, we refer to the term tax recovery to highlight the value of those upward amendments. We also use the term tax returned when either an enquiry or less formal investigation suggests an overpayment of tax, which is then refunded.

Tax recovery has been variable to date which highlights the different nature of the cases we have worked since 2019. In that first year, we were primarily focussed on more obvious cases of tax risk, which were relatively easy to resolve. Having since closed down those risks by amending our system, we have concentrated on different risks with a general stabilisation over the last two years, and around £1.2m recovered in 2022-23.

Tax returned grew a little in 2022-23, mainly due to addressing cases where taxpayers had overpaid tax following the change in tax rates in October 2022. Prior to that tax returned had been quite stable.