Agenda item 3 is on the administration of Scottish income tax. We took evidence on 3 February on the reports that we had received from Audit Scotland and the National Audit Office. We want to explore further some of the implications of the reports, and we have a series of questions on them.
I welcome our witnesses. Alyson Stafford is director general of the Scottish exchequer, Fiona Thom is head of the income tax and reserved taxes unit at the Scottish Government, Jonathan Athow is director general for customer strategy and tax design at Her Majesty’s Revenue and Customs, and Jackie McGeehan is deputy director for income tax policy at HMRC. You are all very welcome.
I invite Alyson Stafford to make a short opening statement. We will then proceed to ask a series of questions.
Thank you, convener, and good morning to the committee. Scotland has clearly been in the vanguard in implementing a rapid devolution of fiscal powers, and the onset of powers in Scotland to vary tax rates and bands for non-savings, non-dividend income tax from 2017-18 has relied on the administration of this partially devolved tax by HMRC.
The areas of assurance for the Scottish Government come from three sources. First, HMRC has a designated and additional accounting officer who is responsible for Scottish income tax. Secondly, the Scottish Government and HMRC have a service level agreement with a performance framework that is vigilantly monitored and adapted as our mutual experience grows. Thirdly, the National Audit Office provides an audit opinion, which has confirmed, among other assurances, that the Scottish income tax outturn for 2019-20 has been fairly stated by HMRC.
I also take comfort from the Auditor General for Scotland being satisfied that the findings and conclusions in the NAO Comptroller and Auditor General’s report are reasonably based. That also provides the Scottish Parliament with valuable assurance with regard to this important aspect of the Scottish budget.
That said, we continue to work with HMRC and Revenue Scotland to ensure that the Scottish public finances are underpinned by tax administration arrangements that function effectively and as intended, ensuring value for money for the taxpayer.
My colleague Fiona Thom from the Scottish Government’s tax directorate and I look forward to answering the committee’s questions.
Thank you for setting out that introductory framework. I am sure that you will have read the evidence that the committee took on 3 February. We were particularly exercised by the fact that we are in years 4 and 5 of a distinctive Scottish income tax system but there appear to be what, in our eyes, look like significant gaps in the data that is available and in the evidence that we think is necessary to allow Parliament and the Government to make informed decisions and choices about income tax policy in Scotland.
We are particularly interested in the level of compliance activity that is Scotland specific, in whether there is sufficient—or any—data on the tax gap in Scotland, and in the extent to which the information can be interrogated and analysed, which we think is extremely important in fashioning the evidence upon which we can build a sustainable and effective tax system.
You mentioned service level agreements, which we will return to. We are very pleased that the Scottish exchequer and HMRC are here as parties to that agreement, because we will want to explore that in our questions.
I call the deputy convener, Sharon Dowey, to open the questioning.
Good morning, everybody. In evidence to the committee, the Comptroller and Auditor General concluded that HMRC’s outturn figures and administration of the system were reasonable, and that the administration of Scottish income tax had
“now reached what is essentially the implementation of business as usual.”
He also said:
“HMRC’s focus must now be on refining its processes to maintain an accurate and complete record of the Scottish taxpayer population and on continuing to monitor the risk of non-compliance that might or might not arise as a result of divergence between UK and Scottish tax rates.”—[Official Report, Public Audit Committee, 3 February 2022; c 3.]
What areas of refinement are still required
“to maintain an accurate and complete record of the Scottish taxpayer population”?
Scottish taxpayer identification is absolutely at the core of our administration of Scottish income tax. However, that is an on-going process; we cannot do it once and then leave. After all, the population is clearly not static. Every year, we carry out a postcode check, while every two years, we check the address data that we hold in our systems against third-party data—electoral rolls, credit agency and post office data and so on—and match individuals with it. We are able to match 70 per cent of those individuals in that exercise. When we look at those matched individuals, we find that, in 99.9 per cent of cases, the address is the same. In a small number of cases, however, the address is different and, for some, it falls on one side of the border or the other, and we then need to check which side of the border—Scotland or England—the individual is living on. We write to them and ask them to update their address records or tell us if their addresses are wrong.
With the remaining 30 per cent for whom we do not get a match, we look for the addresses in the pay-as-you-earn records in our real-time information system or in our self-assessment records. In those cases, we are able to corroborate them. In total, we have corroboration for 98 to 99 per cent of Scottish taxpayers.
With the remaining 1 per cent, the information is not necessarily incorrect—we have just not been able to corroborate things. We have a very good level of accuracy, which we maintain by communicating with taxpayers and encouraging them to tell us when they change their address so that we can keep that information up to date. We also get addresses from other sources, but we maintain that approach of constantly repeating our communication with taxpayers and checking our databases to ensure that the addresses are in the correct format.
Do the errors appear to be intentional or unintentional?
We see no evidence of intentional errors. Most people tell us when they move house, so we get their address then. We sometimes get incorrect postcodes or addresses in a format that does not quite work with the system, and we have a process for checking and correcting postcodes. For people who are in employment or receive a pension, we correct the incorrect postcodes and ensure that they are recorded accurately as Scottish taxpayers so that no revenue is lost to the Scottish Government.
Thank you.
My first question relates to the comment in the NAO report that
“HMRC has limited performance data available about its compliance activities in Scotland.”
Why is that the case, and what, if anything, is being done to try to rectify the situation?
There are two key factors, the first of which is that, as Jackie McGeehan has said, we do not think that there is a particular issue with people not complying or trying to misrepresent their address. That reflects partly the work that we are doing and partly the fact that there is not a large differential between the rates of income tax.
Secondly, we try to take a holistic approach to non-compliance. We consider how we can minimise non-compliance across the United Kingdom as a whole and across all the taxes that we administer. Obviously, income tax is one of the largest revenue raisers, so it requires due care, but we look across all taxes.
Our approach involves looking at how we can maximise compliance activity for the UK as a whole. That compliance activity is not just about chasing money, although we obviously want to collect all the money that is due under the law. We also have anti-money laundering responsibilities, and we sometimes tackle fraud cases that involve organised crime. Our compliance activity has a number of objectives; some are about collecting money and some are about tackling wider illegal activity.
From that point of view, we try to achieve the best return—not necessarily a financial return, but the best public policy return—across the UK as a whole. That is how we tend to operate, which means that our systems are set up to record that. Although there will be different geographic patterns between Scotland and the rest of the UK, or at more granular levels than that, our systems are not set up in that way, because we try to optimise our activities.
You say that the variation is not particularly significant, so you would not have concerns, but someone in Scotland who earns £50,000 will pay £1,489.10 more in tax than someone in the rest of the UK who earns £50,000. Given that there is always compliance, evasion and avoidance, at what level of variance would you start to have significant concerns and think that you would need to probe deeper into the data?
It is very hard to give an exact figure, because it would depend on what patterns we were seeing in the compliance work that Jackie McGeehan talked about. It would depend on whether we were seeing more people with problems with their residence and whether we could identify that as deliberate action, such as deliberately misstating or not updating information, or other negligent behaviours. We would look at those sorts of behaviours, as well as the difference in rates.
Our approach to compliance is different for different taxpayer groups. We have a dedicated team of people who look at high-net-worth individuals, and those in Scotland will be managed differently. High-net-worth individuals might be a group who you think are most at risk of trying to arrange their affairs in a way that minimises their tax, so we have a unit that looks at those individuals, and their residence will be part of that work.
It is a question of taking a tailored approach, not just geographically, but also in relation to the different types of taxpayer that we have and the different opportunities that different taxpayers will have to arrange their affairs, whether legally or less so. That is the approach that we take.
The Scottish Government has taken the decision to raise tax on middle-income earners upwards. You identified high-net-worth individuals or millionaires. Let us take the example of a millionaire with a house in North Berwick and a house in Berwick-upon-Tweed who works in London and Edinburgh. That is a conceivable case. Is that now a grey area that concerns you?
No. We have clear rules on residence. Of course, there have been long-standing issues around residence internationally, as well as within the UK. Jackie McGeehan, do you want to comment on the clarity of the residence rules?
10:15
We think that the residence rules are clear in legislation. For most people, where they live is absolutely clear cut, because they have one residence that is in Scotland, England, Wales or Northern Ireland. For those who have two residences, the question is about where their main residence is and where they spend most of their time—basically, where their life is. If someone spends most of their time in Scotland, they are a Scottish taxpayer.
Where someone has multiple residences and it is not clear where their main residence is, it becomes a day-counting issue to find out how much time they spend in each of those residences, before we reach closure. However, that is fact based. People cannot choose and say, “I fancy being a Scottish taxpayer,” or “I fancy being an English taxpayer.” That is based on where they actually live and spend their life.
If I fly in and out of the country from Buenos Aires, you can probably monitor that and count the days much more easily than if I go between North Berwick and Berwick-upon-Tweed. How do you monitor that? Do you just take what people say at face value and take their word for it?
In most cases, we do not think that there is an issue but, if we carry out an inquiry, we would seek evidence about where the person is living. For example, that would be about where their children go to school or where they are a member of a golf club. We would not often get into that level of inquiry, but there is evidence that we can look at to decide where somebody actually lives.
Thank you.
In our previous session on the issue, Colin Beattie went into some detail on one aspect of the report, and he is keen to get back into that level of detail on that aspect, so I invite him to put some questions.
I have a couple of specific questions. After the previous meeting, I looked at the National Audit Office report and extracted from it just an A4 page—it could have been more, but I kept to the main points—setting out the references to estimates, information not being available, projections and all sorts of other things. If we take each issue individually, perhaps they are explainable but, if we take them in aggregate, surely the impact on Scottish income tax is significant.
I do not know whether you have done a crude exercise such as the one that I have done, but it seems to me that, taking the issues in the round, there must be concerns about the accuracy of the income tax take, which obviously has huge implications for the Scottish Government and for HMRC. In a general sense, how are you going to deal with all those issues? Are we going to get away from all the estimates and the fact that we cannot identify individual figures and so forth? Maybe that is for Jackie McGeehan.
I will come in on that. We are talking about a number of different figures, and I would put them into three groups. First, we think that, of the amount of income tax that is Scottish income tax, 97 per cent is identified through the PAYE system or self-assessment, so it is actual liabilities that have been established. That involves just adding up what residents in Scotland pay through PAYE or self-assessment. In that regard, all that we are doing is adding things up from the forms—we are not making any estimates.
That is one element. There is an amount that we have to adjust, because we know that some people with self-assessment returns will be late filers, so we will not have the information. The 3 per cent is estimated, but 97 per cent is based on established liabilities, so we have a high degree of confidence on that. Although there are estimates, they do not affect the 97 per cent.
There are an awful lot of estimates.
There are. There are two other elements where the estimates are needed. One is in relation to the forecasts. The Office for Budget Responsibility and the Scottish Fiscal Commission make forecasts. Some of that forecast data relies on some of the assumptions that you have in front of you. That does not affect the outturn—the money that is actually achieved—but it is important for setting budgets and overall public spending plans at a UK level and in Scotland. Therefore, those estimates affect the forecasts. There is also the block grant adjustment, which some of those assumptions will affect. However, going back to the outturn data, 97 per cent is based on established liabilities.
Having worked in related areas for a long time, I know that we need to make lots of assumptions. Particularly with issues of tax—we will probably get on to this later—there is always a need to understand what non-compliance is. It is very difficult to tackle that and it is necessary to make assumptions. I come back to the fact that 97 per cent of the outturn tax collected is based on established liabilities. The 3 per cent requires estimation, but 97 per cent is pretty clearly established.
I emphasise the difference between the National Audit Office opinion on the actual tax recorded for 2019-20 and the provisional figures that have been audited for 2021. As Jonathan Athow said, the 2019-20 figure is the one that we have to rely on as the tax actually collected and due to Scotland. That is a really important figure for us in Scotland. It is the one that we use to assess the performance of the forecasts that we used in setting the budget, which come from the Scottish Fiscal Commission.
The difference between the Scottish Fiscal Commission’s forecast and the block grant adjustments that are associated with it was £34 million. That adjustment had to be made in our current year—the 2022-23 budget. We rely on those elements of estimation to arrive at the figure of £11,833 million. Reading the detail of the National Audit Office’s opinion and the assurance that it can give the Government on that is essential.
I do not wish to diminish the work that the National Audit Office has done on the provisional figure for the 2020-21 year, but we do not use that. We rely on the forecasts that the Scottish Fiscal Commission generates to set our budget and expenditure plans and to determine the flow of cash from the Treasury in that year.
The estimate on which I am most vigilant in order to get from the National Audit Office’s opinion as robust a position as possible on HMRC’s work is the 3 per cent that Jonathan Athow mentioned. The impact of the net adjustments was £150 million, which would be a net estimation of -1.3 per cent.
It is important to make the distinction about which numbers we actually use in Scotland. The number that the National Audit Office is considering for 2020-21 is for the purposes of the HMRC account. We have to wait until the full outturn, when all the self-assessments come in, to get the final figure for 2020-21.
I wanted to clarify the situation and say where the emphasis of assurance is needed for us in Scotland. We use the outturn figure and we rely on the Scottish Fiscal Commission for all the provisional work.
There is a fairly long list of estimates and workarounds in the NAO report. Do most of those exist for income tax in the rest of the UK or are they specific to Scotland, because of the way in which the settlement has been done?
A large number of estimates are also used at UK level. As we have already explored in relation to compliance data, it is more challenging to produce Scotland-specific figures for some estimates because of how the data is collected and because of historical decisions about the set-up of information technology systems. We do not always have a split with Scotland, Wales, Northern Ireland or regions of England. We often do not have data at lower levels of geography that would allow us to break down some of the assumptions. Assumptions are used at UK level, and we sometimes apply a UK assumption to Scotland.
Given that there is devolution across the different nations, you would expect that some effective work would be taking place so that individual figures could be given for individual nations.
As I said, we can be very specific about certain estimates and figures. I go back to what I said in relation to the 97 per cent figure. The challenge is whether we have the data available for some of the other estimates. I know that a wider conversation is taking place between the Scottish Government and the Scottish Parliament about what other economic data should be available for Scotland. Economic data is often needed to work out the share of an activity that arises in Scotland. There is a wider programme of work to provide more data, but we are sometimes constrained by decisions that were made well before many of the current devolution arrangements were ever considered. That limits our ability to provide more granular estimates.
I remain of the opinion that the aggregate totals for all the estimates, workarounds and so on must be significant, which is a concern.
I will move on to the specific issue of missing Scottish postcodes, which you have touched on. My concern is that the number had increased from 13,708 to 23,351 when the NAO last reported. That is a 70 per cent increase, although the point has been made that the figure represents something like 1 per cent of the population. However, the concern is that those missing taxpayers can equate to a fairly large sum of money in tax, especially if they turn out to be individuals with a high net worth.
I have three questions. What is behind the increase in the number of missing Scottish postcodes? What is the impact on revenues likely to be? What is HMRC doing to fix the issue?
I will start, and then I will hand over to Jackie McGeehan. We want to monitor the issue of missing postcodes, which is an important indicator for us. As I said in my earlier answers, we want to make certain that we understand what is happening with that number and what the longer-term trends are. Although four or five years might feel like quite a long time, we need a bit of a track record when operating tax systems in order to understand how data changes over time.
As Jackie McGeehan said, if we cannot match a postcode, or if we have problems with postcodes, we have a process for working through that. Sometimes, we are able to identify that the taxpayer no longer has any liabilities. We might no longer have any record or postcode for someone because they have left work, which might mean that they no longer have a tax liability. There will be reasons for that, and what we will do is work through all those postcodes and identify where we think that a missing postcode should be in Scotland. So far, we do not see any widespread problems here. We will want to monitor how that number changes over time, as it is an important indicator.
10:30To go back to your point, there are 2.5 million income tax payers in Scotland, so we should remember that, when we talk about a figure of 25,000, although that seems quite large, it represents 0.1 per cent of taxpayers, which means that the numbers that we are talking about remain small. We do not think that there is any disproportionate effect on the income tax that is raised in Scotland.
The address data that we get comes from a number of sources. Individual taxpayers tell us their address, employers can give us the address of employees who join them and we get data from other Government departments. As you might imagine, as is the case with many other organisations, the data that we get is not always perfect, and we need to do quite a lot of work to cleanse it and make sure that it is right. We check those postcodes annually and we correct all the postcodes for people who are in employment or in receipt of a pension and are taxpayers. We correct all of those in year, so there is no loss to the Scottish Government in terms of the revenue.
As Jonathan Athow says, we will monitor the situation, but the data comes from other sources and we put it right to the extent that we need to in order to get the right number in terms of tax for Scotland. However, it is important to remember that we are talking about a tiny proportion of the tax that is due to Scotland.
Jonathan Athow, you said that the figure that you gave represented 0.1 per cent of taxpayers. Is it not 1 per cent?
Yes—sorry, that was my fault. It is 1 per cent.
I just wanted to be sure that we are on the same page.
You have not said what is behind the increase.
At the moment, we do not have a good indication of what is behind the increase. We need to understand what is going on. Some of the issues might arise because of changes in the economy or they might be to do with people moving between the different categories that Jackie McGeehan talked about, which might be creating more noise in the system. At the moment, we do not have a good indication of that. Is it to do with economic change or is something else going on? We will continue to monitor the situation, because we think that it is a valuable piece of information that might give us an early indication of other concerns that we ought to have about the way in which the system is operating.
You did not indicate the scale of any impact on revenue.
As Jackie McGeehan said, we will work through that. We do not think that there is any meaningful impact on revenues. What we are dealing with is the starting position in relation to those taxpayers, and we will do all the work to ensure that we have a proper postcode for those cases in which there is a problem. We think that, by the end of it, there will be no residual problem. We are confident that we are able to tackle the issue but, obviously, we want to keep an eye on the situation to ensure that it remains a small problem.
Westminster’s Public Accounts Committee has reported on HMRC’s management of tax debt, particularly in relation to the Covid period. The committee came up with a number of conclusions that it set out in six bullet points, some of which are quite strongly worded. It says:
“We are not satisfied that HMRC has a clear plan to tackle the mountain of tax debt which has built up during the pandemic ... HMRC is not being ambitious enough in bringing down debt levels and securing the resources this will require ... Rogue companies are exploiting the pandemic to profit at the expense of taxpayers.”
Scotland will not be exempt from that last issue, I am sure. The report continues:
“HMRC is far behind where it needs to be in making good use of data to manage debt effectively ... HMRC is not using all relevant data sources to understand how the pandemic is affecting taxpayer’s ability to repay.”
Finally, it says:
“We are concerned that HMRC is not doing enough to identify vulnerable people who need extra support with their debts.”
Has HMRC accepted those conclusions?
I do not think that we would express the issue in quite the same way. During the pandemic—
I am sorry—what does that mean?
Elements of the report bring out some challenges that we face, but I would not characterise the situation by saying that we do not have plans or approaches, although our plans and approaches may need to change.
Therefore, HMRC has not accepted the findings of the committee report.
I would not say that. Let me go back to what I said before—
Is it not a yes or no answer?
As I said, elements of the report are absolutely true, but I do not think that the way that it is characterised is correct. We will probably partially accept elements of the report, because there is a great deal of important work to be done there.
To go back to the overall issues, during the pandemic, we deliberately chose to make changes in the way that we collected debt, because we recognised that businesses and individuals would be in financial distress. That caused a very large rise in debt—our outstanding debt balance rose to £70 billion. We have brought down that balance and, at the end of the previous financial year, there was £40 billion of outstanding debt. However, there is further to go, and the report says that we now need to work through how we continue to get the debt to fall. We have already made substantial inroads into that stock of debt, but there is further to go.
In that respect, the Public Accounts Committee is absolutely right that there is more to be done to reduce debt back towards the pre-pandemic level—I am completely with the committee on that. However, I think that the way that the committee’s recommendations are expressed implies that we have not done anything or that we do not have clear plans. We are working through those plans, and whether those plans will need to change is another question. This has been an unprecedented time for us, and we need to work out what we need to do differently now.
Has HMRC responded to the committee report?
To the best of my knowledge, we have not responded yet.
You have a huge task because, before the pandemic, you managed about 3.8 million taxpayers in debt and, as of September 2021, that number was 6.2 million. That is a huge hill to climb.
Indeed. As I said, the overall stock of debt reached a peak of £70 billion during the pandemic. We have brought that down to around £40 billion, but that is still above the pre-pandemic levels, and the figure that you mentioned is not entirely consistent with those levels, so there is still further to go.
The challenge is to make certain that we strike the right balance. Some individuals or businesses will still be in financial difficulty, and we wish to support those taxpayers who cannot pay. However, it is also important that we collect money from taxpayers who can pay because, as you all know, that money is important for the UK and Scottish Governments to fund public services.
Do we have any specific figures on that for Scotland?
I do not know but, if we have anything on Scotland, I can update you.
I think that this committee would be interested if there are specific figures about taxpayers in debt in Scotland.
I will take that away and write to the committee about that.
Could you tell us when you will officially respond to that Public Accounts Committee report?
As I understand it, there is a three-month grace period for that, so we would normally respond within that time. I cannot comment any more than that. It is not fully my area of responsibility, so I am not fully sighted on all the ins and outs of that report.
Okay. Obviously, we will await that response with interest.
I want to go back to something that Alyson Stafford said. As I understand the process, amendments to the Scotland Act 1998 provided for the setting of a Scottish rate of income tax from April 2017 onwards. HMRC collects and administers Scottish income tax, HMRC’s accounts are audited by the National Audit Office, and the Comptroller and Auditor General is required to report to the Scottish Parliament on HMRC’s administration of Scottish income tax. His seventh report on Scottish income tax was laid in Parliament on 14 January 2022. However, you told us earlier that you do not take any account of the NAO’s estimates of Scottish income tax. Why do you ignore that important body of evidence and rely solely on the Scottish Fiscal Commission’s estimates?
The fiscal framework, which is the arrangement that has been agreed between the Scottish and UK Governments, determines how we set our budgets in the first place, and how those budgets are adjusted when we have the outturn data. Those arrangements require us—this is in legislation—to use the forecasts as generated by the Scottish Fiscal Commission when taking into account the level of income tax that we bring into our budget each year.
That is the arrangement that is in place, and we are absolutely complying with that. The Scottish Fiscal Commission sets out and publishes how it generates the forecast for the partially devolved taxes such as income tax, but also for taxes that are fully devolved—those that are collected and administered by Revenue Scotland—and the non-domestic rates income, which is in effect a tax that is administered by local government.
We use what we are required to use under the fiscal framework and the legislation in Scotland. Obviously, the Scottish Fiscal Commission has access to the real-time information that is now available and it draws on different data sources—it publishes the basis on which it works through its forecasts. As I say, however, the strict requirements for what we need to use are set out in the fiscal framework and we use the Scottish Fiscal Commission’s forecasts as set out in Scottish legislation.
I understand that you are compliant with the requirements that are placed on you by legislation but, as the director general of the Scottish exchequer, would it not make sense for you to at least take into account the National Audit Office’s estimates of future tax take?
What makes sense for me is to use what comes from the Scottish Fiscal Commission.
Does it use the NAO’s estimate as part of its deliberations?
It uses a range of data sources, and they are published.
Does that include the NAO’s estimate?
It uses a range of things from different sources, although not necessarily the NAO. Some of it will be from HMRC, and the commission will use other sources. Those are the figures that we use. We are required to use those in operating the fiscal framework in Scotland as part of the whole package of fiscal devolution.
Okay. Maybe as DG for the Scottish exchequer, you could write to let the committee know what are the sources that are used.
Do you mean by the Scottish Fiscal Commission?
Yes.
By all means, we will provide the link for all the different data sources. That information is published, but we are happy to do that.
Thank you, but we are still not clear about whether the NAO is included.
I might be able to clarify that. As far as I am aware, the Scottish Fiscal Commission does not use the HMRC provisional estimate as part of its current forecasting data sources. Obviously, if you want more information about why that is the case, you would have to speak to the Scottish Fiscal Commission but, as far as we are aware, the commission does not use that estimate in its current forecasting of Scottish income tax.
That figure is, however, an additional Scotland-specific number that we can compare against what we get from the Scottish Fiscal Commission and, if those figures were completely different, we would be aware of that and would discuss it with the Scottish Fiscal Commission. However, to answer your initial question, we do not think that it uses the NAO estimate, but we can send a list of data sources.
I am not trying to trip anybody up, but are you saying that you do not think that the Scottish Fiscal Commission uses that estimate, but you use it as a reference point?
No. Alyson Stafford has stated clearly that we do not use it, because we use the Scottish Fiscal Commission’s forecast for our budget. However, it is an additional piece of data and it is not the case that we never look at it. It is part of our overarching understanding of Scottish income tax, but it is not used when we set our budget, because that is done using SFC figures and the actual outturn data.
10:45
Okay. We might return to that before we finish.
My next question is for Jonathan Athow. The Audit Scotland report that accompanies the NAO report made a point about good governance and assurance arrangements and the need to keep those under continual review. It said that the process should include
“ongoing consideration of the frequency of third-party data checks and Service Level Agreement performance measures, such as setting compliance target levels for Scottish taxpayers without ‘S’ prefixes.”
Why has HMRC still not introduced a target in relation to the number of missing S prefixes? Will it do so at some point in the future?
I ask Jackie McGeehan to explain briefly how S prefixes work, after which I can talk about what we will do in that regard.
When we identify a taxpayer as Scottish, we send their employer their tax code. Every year, we send out tax codes to the employers of all pay-as-you-earn taxpayers. We usually do that in advance of the start of the tax year. That is the code that they should apply to that individual. In cases in which they are Scottish, the code has an S prefix.
Those taxpayers are correctly identified by us as Scottish. In a small number of cases, the employer—for whatever reason—does not apply the code correctly. In other words, they do not treat the individual as a Scottish taxpayer. They do not change the code; they apply a different code. We identify those cases and take action. We write to the employer and say, “You haven’t applied the correct code. You need to apply it—here it is. Please do it.” In cases in which employers fail persistently, we contact them to find out why. Sometimes, they have problems with the software; sometimes, they simply take a long time to update the codes. We have a programme of educating those employers and supporting them to get it right.
When that happens, such taxpayers are still correctly identified as Scottish on our systems. At the end of every tax year, we do a reconciliation of our pay-as-you-earn scheme system, whereby we look at what the employers have deducted and at what we would expect them to have deducted, based on our codes and what we think they should be paying. If there is a difference—if someone has overpaid or underpaid—we will correct that.
Therefore, the Scottish Government is getting the tax that it should get. Ultimately, the individual pays the right amount of tax. Unfortunately, when an employer gets it wrong, with the result that the individual has the wrong amount of tax deducted during the year, they may find that they have a bit more to pay. Those are employer errors. We take steps to reduce that number of errors as far as we can. I think that around 4 per cent were incorrect when we first started issuing S codes, and the figure is now down to around 1 per cent. The number is coming down, and we will push to make sure that it stays down and goes down further.
The main thing to draw from that is that, in the end, the right amount of tax is collected. There is an inconvenience here. Therefore, when it comes to the operation of S codes, it is not really a question of money not being collected; the problem is with the good governance and good operation of the PAYE system. There is a wider small problem that we have, which involves some PAYE systems not being operated correctly. We work with employers to make certain that they are collecting the right amount of tax. However, that is a very small issue.
As far as the S codes issue is concerned, we do not think that that is a particular risk to the collection of the right amount of tax. When it comes to data sharing more generally and how the service level agreement will work, we already do quite a lot of data matching. The challenge is to make certain that we respond as events develop. I have already talked about the number of postcodes that are not correctly identified. As the process becomes more mature and new issues arise, we will have to respond. I think that we have the tools to do that, whether we are talking about postcodes, S codes or anything else.
At the moment, we are confident about the system, but we cannot be complacent; issues might arise as the system develops, perhaps as there is greater divergence in tax rates.
Okay—thank you. I am conscious of the time.
We have mentioned in passing the impact of the pandemic. Willie Coffey has a series of questions to ask on that.
The conversation earlier was fascinating. What I drew from it is that we can have any number of forecasts, but it is the outturn that matters. Alyson Stafford led the discussion about the arrangement being firmly embedded in the fiscal framework. I am sure that some people are asking about revisiting that if there is any scope to do so. Who knows? That might evolve in the future.
I want to ask a couple of questions about the impact of the pandemic on tax receipts. I also have a little question about the compliance issue that is mentioned in the report.
In a meeting earlier in the year, we heard that the forecasts for income tax receipts had increased significantly. Will you tell us a little about your estimation of the impact of Covid on your forecasts? Have those forecasts changed as the result of the impact of Covid? That is probably a question for Jonathan Athow and Alyson Stafford.
I will talk about the high-level impact. There have been a number of effects, one of which has been an impact on operations. I have already talked about debt. We are talking about large numbers in that regard, but things are a bit uncertain. There have been impacts on the amount of cash that is coming into the Exchequer.
On underlying liabilities, one thing that we have seen is that the impact of Covid on receipts has perhaps been less pronounced than in many of the earlier forecasts. That is because of the overall labour market position, and particularly that on income tax. Employment is a key driver of that, and that has held up better than people expected. The levels of employment have not got back to pre-pandemic levels, but they have recovered quite quickly. We have also seen wage growth holding up quite well. Obviously, for tax receipts, the number of people in work and the wages that are paid—and growth in both of those—are important.
The situation has perhaps been more encouraging than we might have thought that it would be. Obviously, self-assessment money is paid with a long lag, but the self-assessment receipts that came in in January and February this year, which covered a pandemic year, were higher than forecast. That suggests that the impact of the pandemic for self-employed people was not as large as was first thought.
As I have said, there are still question marks. We do not know how many companies or individuals are still struggling with a legacy of debt or other such issues. There are still a number of uncertainties, and a long tail of companies might struggle as the economy develops. In addition, there is a whole set of new challenges with rising energy prices and supply chain disruptions in China and from the war in Ukraine.
There are lots of uncertainties but, in general, receipts have held up better than most people expected them to.
What is Alyson Stafford’s view on that? Do you mirror those estimates? Are you similarly confident that there is a possibility of a better outturn than was first anticipated?
I will take that in two parts. In relation to 2019-20, we have the NAO’s assurance on actual outturn, but those figures were predominantly pre-pandemic. Therefore, we have not really seen any pandemic impact on the figures for that year.
On what that means for our budget, as we have discussed, the Scottish Fiscal Commission sets out a forecast position. That is the figure that we used in our 2020-21 budget. The outturn on that will be published by HMRC this summer, so we will see that figure then.
What data do we have in the meantime? I recognise a number of things that Jonathan Athow is explaining from our economic analysis. I would also point out that, in the real-time information that we were getting from HMRC, which comes from the PAYE system, we did not see any particular shock in that flow at that time. There is a lag in getting that data, and there are still things in it that must be refined and cleaned, but that at least provides some source of data in those particular years.
I await with interest the actual final outturn for 2020-21, and any reconciliations that are needed will be applied in the budget for 2023-24. There is an interesting gap between seeing how the economy performs and then seeing how that performance filters through into our public finances.
Is it too early to say anything about the possible impact of the various support schemes on your forecasts?
At the moment, we are not seeing any particular effects. We did not see any pronounced effects from the ending of the coronavirus job retention scheme—the furlough scheme—last autumn such as, say, people falling out of the workforce. I am certain that there were some unfortunate people who lost their jobs as those schemes wound down, but there were no effects large enough to affect our forecast. Please do not interpret that as me saying that Covid had no effect—there has been a dramatic effect in receipts, However, the extent of that is perhaps less than people first thought that it would be.
I just want to add that we usually publish the medium-term financial strategy at the end of May—it is one of the points in our fiscal calendar—and there will be more information on tax revenues and relative earnings growth over the past few years. Some of the analysis that we have will be available to Parliament then.
Finally, according to our papers, you have opened 29 per cent fewer civil compliance cases, which is much fewer than before. Can you say a bit more about the possible impact of that on tax revenues?
Again, the pandemic meant that we as an organisation had to radically change how we worked. We had to implement a number of Government support schemes from scratch in a matter of weeks, which meant that we had to redeploy resources and divert some of our activity away from compliance matters. That was a conscious decision at the time.
The question that we are grappling with and to which I do not think that we have the full answer yet is the extent to which those things have been deferred and will be caught up with, and the extent to which the bus has gone and we have been left at the bus stop as far as those cases are concerned. We are continuing to work through that.
We have plans to increase our activity to normal levels as we come out of the pandemic, but it is still unclear whether there has been any permanent effect on the amount of compliance revenue that we can collect. As I have said, we are trying to work through that, but the pandemic more generally has left us with many unknowns with regard to what will happen in the coming years. Debt is another example of where we are starting to work through all the implications but have not done so fully yet.
Have you been able to put any estimate on that at all?
Broadly, we think that we can make back much of that. After all, you start a compliance investigation because you have evidence of something wrong. Those cases can still be worked.
It might be a challenge if a taxpayer has disappeared or has become insolvent, because those cases cannot be fully worked through. In general, though, we are confident that we should be able to make back a large part of that dip in compliance activity in future. Again, there are uncertainties around that.
But you do not have a figure.
I would have to write to you on that, but even now any figures will be estimates of what we intend to do in future rather than firm numbers for the money that we have brought in.
I would be obliged if you could submit that to the committee.
11:00
I will pick up on that contemporaneous point before I come to my final question. Around the time of the UK spring budget statement, the Office for Budget Responsibility produced an assessment in which it suggested that there might be some buoyancy—a euphemism for an increase—in tax receipts, as a result of the rise in inflation. I presume that that assessment was based on wage and salary demands and rises perhaps going up at a higher rate than they have been for the past decade as a result of the fuelling impact of price inflation. My question is first to Jonathan Athow. Are you seeing any impacts from that rising inflation in your tax collection levels?
You are absolutely spot on in the sense that there is that combination of wage and price rises. VAT is charged on the purchase price and, if that purchase price goes up, we get more tax. Therefore, we are seeing some buoyancy.
With income tax, the issue is not just average wages but the distribution of wage growth. If there is more wage growth at the top end of the labour market among the very well paid, where the marginal tax rate is 46 per cent here in Scotland or 45 per cent in the rest of the UK, that obviously brings in more tax receipts than if the wage growth is for basic rate taxpayers. Some of what we have seen, including evidence that has emerged more recently, suggests that some of the buoyancy is because of wage growth among higher-paid people with a higher marginal tax rate.
You are right that inflation tends to boost tax receipts, although there are many other factors at play.
My next question is for Alyson Stafford and Fiona Thom. Figure 9 in the NAO report shows where there is most divergence in tax paid by earnings. Those who are on around £50,000 in Scotland pay higher income tax than those in other parts of the UK. In light of that and of what Jonathan Athow has just said, is the Scottish exchequer monitoring the situation? Do you have any sense that tax receipts might be going up, and do you have any plans for how they might be spent if they go up?
I will ask Fiona Thom to come in on the specifics of how the inflation factor works in relation to the fiscal framework. Let me be clear that it would not be about whether I had any plans to spend anything; it would be for the Cabinet or ministers to decide. However, before we all get too excited, I will bring in Fiona Thom.
I echo Jonathan Athow’s point. For Scottish income tax receipts, Jonathan described one of the effects that inflation has had and that the OBR forecast in its spring statement. The other effect that inflation can have is if bands and rates are indexed by inflation. In the rest of the UK, bands and rates are frozen, so the personal allowance and higher rate threshold are frozen. In our system, we have uprated some of the bands by inflation, which will reduce some of the tax take. There are two kinds of effects, and whichever is the strongest will come through into the overall effect, so we might not see an increase such as the OBR has forecast.
Additionally, we know that fewer taxpayers in Scotland are at the top end of the distribution, which is one of the challenges that we face with the net position for Scotland. We have to wait and see what the Scottish Fiscal Commission publishes in its next forecast, which, as Alyson Stafford said, will be alongside the medium-term financial strategy on 31 May. The commission will consider all those factors.
My final question is on a much more mundane and less exciting issue, but one that is very important to us: the service level agreement between the Scottish Government and HMRC. You will have seen that, in the evidence session on 3 February, we explored with Gareth Davies and Stephen Boyle how the agreement currently works and whether it could be improved. I suppose that my question is really about what consideration has been given to the existing terms of the service level agreement and whether it can be improved and cover things such as the tax gap in Scotland, for example.
The service level agreement for 2022 is currently under review, so the committee’s consideration is extremely timely for us to take your observations and comments into account. The things that we are looking at and having further discussions with HMRC colleagues on are very much in the data space. We will still see the vigilance on data quality that has been the subject of quite a lot of your scrutiny today, but we are also looking at data breadth and depth so that we will be better able to understand the dynamic of the tax base, and at the timeliness of data.
We absolutely recognise the comments in Audit Scotland’s report on the Scotland-specific compliance data, and we are discussing that further with HMRC. Jonathan Athow might wish to comment more—depending on the committee’s time—on the feasibility of that. Those are the sorts of things that are essential for us to maximise the fair and proper collection of tax so that we can apply it to public expenditure in Scotland.
You can comment without prejudice, if you like, Jonathan.
We are very open to reviewing the arrangements. We have a commitment to make this work in a way that works for the Scottish Government and for the Scottish Parliament’s oversight, so we are very open minded on it. There are challenges around some of the data. Sometimes the challenge is not due to a lack of willingness but is about what is practical with the information that we have. However, the position is not static, and what is available today may not be the position on what we can do tomorrow.
One of our questions and interests is whether it is the case that the data does not exist or is impenetrable, or whether we could get the wider data set that we think is necessary if the Scottish Government paid more than £700,000 a year for it.
There are obviously some things that can be done, but some of this is much more ingrained in the system. I will give an example. Sometimes what we do to tackle non-compliance, which will affect Scotland, is not about knocking on people’s doors, whether in Scotland or in the rest of the UK, but is about making system-wide changes. We are putting in place arrangements whereby self-assessment taxpayers will have to use software to report to us. We think that that will reduce errors. That is an activity that we are undertaking, but it is an activity that will affect all of the UK. I do not know how meaningful it would be if, as an additional activity, we tried to assign different elements of that to different parts of the UK. Sometimes we have to stand back and ask what we are trying to get at, and not just break down the data for the sake of it.
It is a complex picture, but that reflects the complexity of administering the tax system.
Finally, do you have a timeframe for when you expect your discussions and negotiations on the service level agreement to be concluded?
That will certainly be during the summer.
You think that they will be concluded by September.
Yes.
Okay. Thank you.
On that note of clarity, I thank our witnesses Jackie McGeehan, Jonathan Athow, Alyson Stafford and Fiona Thom for the evidence that they have given. We appreciate their time and their contributions.
We will consider what steps to take and how we can keep a monitoring eye on this important work in future. As we have said, we want it, above all, to be evidence led. Whether it is around people’s behavioural patterns, compliance rates, collection rates or people fleeing the country in order to evade or avoid tax, those things are important to us, so I thank you very much for your openness in discussing them with us this morning.
11:09 Meeting continued in private until 11:35.