09:30
Our principal agenda item this morning is consideration of the National Audit Office’s report into the administration of Scottish income tax and of the Auditor General for Scotland’s letter of assurance that accompanies the report.
I am very pleased to welcome our witnesses. We are joined in the committee room by Stephen Boyle, the Auditor General for Scotland—good morning, Auditor General. Alongside Mr Boyle is Richard Robinson, who is a senior manager at Audit Scotland. I am very pleased to welcome from the National Audit Office—online, this year—Gareth Davies, who is the Comptroller and Auditor General at the NAO. Good morning, Mr Davies. Alongside Gareth Davies is Peter Morland, who is the director of financial audit at the NAO.
We are quite limited for time, because Mr Davies and Mr Morland have to leave for a prior engagement at around 10.15. We will all need to be disciplined in our questions and answers. Before we get to those questions, I invite the Auditor General to make a short opening statement, after which I will invite Mr Davies to do the same.
Many thanks, convener, and good morning to the committee. As you know, Scottish income tax is a key part of the Scottish Government’s financial powers and an important contributor to its financial sustainability. In my November 2025 report on financial sustainability and devolved taxes, I set out my view on the use of devolved tax powers, including the management of, and risk and opportunities associated with, Scottish income tax.
Today’s evidence session focuses on the administration of Scottish income tax for 2024-25. This is the seventh year that His Majesty’s Revenue and Customs has published Scottish income tax outturns. The 2023-24 outturn is reconciled through the 2026-27 Scottish budget, resulting in an increase of £406 million. HMRC’s accounts also include an estimate for 2024-25, but that does not yet affect the Scottish budget.
HMRC administers Scottish income tax within the United Kingdom tax system. The National Audit Office audits HMRC’s accounts, and the Comptroller and Auditor General reports to the Scottish Parliament on that work today. I have provided the committee with additional assurance on the NAO’s audit work. I am pleased to advise the committee that I am satisfied that the NAO’s audit approach was reasonable and covered the key audit risks, and that the findings in the C and AG’s report are reasonable and reasonably based. The C and AG has concluded that the outturn on Scottish income tax is therefore fairly stated.
I am very much looking forward to this morning’s session, and I am happy to leave it at that point.
Thank you, Auditor General. I invite the Comptroller and Auditor General, Gareth Davies, to make an opening statement.
Good morning. I apologise for not being there in person for the first time in many years. A clash of dates made it impossible, I am afraid. I will keep my remarks short, convener, in the interest of time.
The Auditor General for Scotland has summarised the headline findings about the reasonableness of the estimate and outturn. The approach that was taken by HMRC in calculating the outturn has remained broadly the same as last year, and I have concluded that it remains reasonable. Both the outturn and the estimate demonstrate growth in the Scottish tax take, reflecting, among other things, the increase in income compared with thresholds, as well as the increase in the higher and top rates and the introduction of the advanced rate in Scotland.
Scottish income tax policy continues to diverge from that in the rest of the UK. For this year’s report, HMRC continues to assess the risk of non-compliance as remaining low, despite the divergence. However, given the time lag between the changes to tax policy and information on non-compliance, it is important that HMRC gets on the front foot to check that it has a well-planned approach in place that clearly addresses any change in compliance risks. That is why, this year, I have recommended that HMRC complete its planned evaluation of the Scottish compliance approach as soon as possible, so that it can identify priority areas for improvement.
I have also encouraged HMRC to consider whether there are any more areas for immediate improvement—for example, areas where it could introduce the collection of more Scotland-specific data. We recognise that any changes that HMRC and the Scottish Government agree to make must be practical and cost effective, with a clear purpose and benefit. We would encourage them to fully explore what improvements can be made.
My team and I worked closely with the Auditor General for Scotland and colleagues at Audit Scotland in the audit. I am grateful, as always, for their co-operation.
Thank you very much indeed, Comptroller and Auditor General. We will touch on all of those areas in the course of this morning.
I will begin with the Auditor General. You mentioned your report from November last year, “Financial sustainability and taxes”, which looked in some depth at the dynamics of Scottish income tax setting and the benefits that it brings. You said that, in the 2025-26 tax year, although £1.7 billion will be raised from the Scottish income tax through the policy choices made by the Scottish Government, the impact on the budget will be a net rise of just £616 million, due to what you describe as the “tax base performance gap”. Could you elaborate a bit on that? Why does that gap exist? Is it to do with earnings levels and employment levels? What lies behind that figure?
Yes, I am happy to do so, convener. The factors that you mentioned are relevant. The gap is also bound up in the fiscal framework between the Scottish and UK Governments. In giving evidence to the committee in December, we made a number of recommendations, including that the issue be set out more transparently in some of the Scottish Government’s documentation for its proposed draft budget. Correspondence received by the committee from the former director general for the Scottish exchequer also made reference to some of the structural economic factors behind the gap—including the types of jobs that exist in Scotland relative to other parts of the UK—which are part of the arrangement for how the fiscal framework operates.
In order to close that gap, the alignment between tax and economic policies has to become clearer. That was one of our recommendations. The driver for that—the primary lever with which to close that tax performance gap—would be the prioritisation of higher-paid jobs as part of Scotland’s economic development. The aspiration would be that, in due course, there would not be as significant a gap as currently exists between those two figures.
Richard Robinson might want to say a little bit more about that. Before I pass on to him, I will reiterate that all the money that is set in the Scottish budget and income tax flows through; the difference lies in the spending power in the budget.
Is that due to the block grant adjustment?
That is correct. It is due to adjustments through the fiscal framework.
Richard, do you want to add anything?
Yes, briefly. Today, we are talking about the assurance over those outturn figures, which are driving the difference between the amount of income tax raised and the BGA, as well as any reconciliations that cover that difference. There will be a difference between the amount that flows in from income tax and the block grant adjustment. That may be different from what the Scottish Fiscal Commission calculates as the total tax that could be raised if there was no difference in anything else apart from policy. However, we know that that is not the reality. We know from the report that we produced last year that there are differences in the sectoral make-up of Scotland compared to the rest of the UK—differences in earnings growth, employment growth and some elements of behavioural change. The tax performance gap enables us to see how big those differences are. It could be a useful tool for the Scottish Government in exploring how to go about correcting that, whether through the economy, tax policy or whatever it may be.
In the interest of time, I will ask one final question before I move on. It is about transparency. You mentioned in your previous answer that the Scottish Government could be more transparent about the gap between the money that it raises and how that ricochets through the block grant adjustment as a result of the Scottish fiscal framework. Could you say a bit more about what you expect the Scottish Government to do in order to fulfil your demand for more transparency?
I will refer to a couple of points. I remind the committee that one of the statistics in our November report drew on the Scottish Government’s survey of the public’s understanding of tax, which said that 50 per cent of people in Scotland do not have a good understanding of how devolved taxes operate. There is an opportunity to close that gap. Therefore, transparency is also about helping people to understand and have an awareness of what goes on.
In that report, we also broadened out to talk about how some of the documentation in the budget materials is presented. In the 2025-26 budget papers, the amount that was to be raised through tax was much more clearly set out than what the benefit to the Scottish budget would be, the figure for which was in one of the appendices. We were clear that it would be much more transparent if those two numbers were side by side in the budget documentation.
There has been some small progress in the budget documentation for 2026-27, but it is not yet at a level where those two figures can be clearly seen without having to go to a bit of trouble. I turn to Richard Robinson, because he has been looking closely into that aspect.
The Auditor General summed it up nicely. There is acknowledgement in the budget documents that, as we talked about in our report, there are other factors at play that will affect the Scottish Government’s budget. However, a separate document still needs to be referred to in order to get to the spending power figure, as opposed to the figure for only the headline tax policy, which we also referred to in last year’s paper.
We—or, perhaps, the committee in the next session—will keep that area under active review, because it is important.
My first line of questioning is more for the NAO. Over the past few years, as the Scottish rates of income tax and the Scottish tax system have moved a bit away from the UK rates and system, there has been concern about taxpayers’ behavioural responses. The most up-to-date information that I have is that there is no clear data to indicate that there have been any behavioural changes in response to those moves. Are you aware of any plans for HMRC to update its research on taxpayers’ behavioural responses to income tax divergence? It was previously indicated that such research could become an annual study, which would arguably be more relevant today, given the increased policy divergence on income tax.
I think that you might be on mute.
Thanks. That point lies behind the recommendation in our report. I agree that the message that HMRC has consistently given to us and the Scottish Government is that, so far, it is not seeing evidence of changes in taxpayers’ behaviour in response to the divergence of the two systems.
Of course, if you do not look for something, you will not find it. HRMC is looking for evidence—I am not suggesting that it is not—but it is planning a more comprehensive evaluation exercise into how the risk is changing. That exercise will also consider whether there are procedures in place for detecting non-compliance—in this case, by mis-stating addresses—or changing migration patterns. It will ask whether there is any evidence that will change our current view that there is not a significant difference in risk between the two countries. We recommend that HMRC complete that exercise as soon as possible; we are not completely clear about what its timetable is.
I agree that the greater the divergence, the greater the risk that it takes too long to pick up on changes in behaviour.
You mentioned only two issues in connection with behavioural change, one of which is providing misleading addresses. However, what about simple behavioural changes that people can make to their lifestyles and the way that they invest, or any other changes that are necessary to reduce their tax burden—I assume legally? Are there indications about that kind of behavioural change, and how quickly would we pick those up?
09:45
I think that there are two—[Inaudible.] Everybody is motivated to pay the right amount of tax but no more. HMRC invests significant effort in understanding taxpayer behaviour across the UK.
My answers so far have been focused on the differences that might be emerging in Scotland compared to the rest of the UK. A significant body of work is under way, which we audit separately from the report that the committee is discussing, to identify what is happening to the tax gap—in other words, whether there is evidence that the amount that is being collected in comparison to what should have been collected is changing over time and whether there are any changes in behaviour. The ways that people try to avoid tax are many and various, and that is a significant part of HMRC’s work. All of that applies to Scotland as well as to the rest of the UK.
The specific question that our report tackles is about the divergence between the two sets of arrangements changing the risk, particularly around the border, in relation to the mis-statement of addresses. You are right that the broader question of taxpayer behaviour is the subject of a large amount of HMRC’s work.
I presume that that applies across the UK. It is a strong indication as to where tax revenues are heading.
Exactly. In the past 12 months, the NAO has prepared a detailed report on HMRC’s approach to ensuring that wealthy individuals are paying the right amount of tax. That is a key focus for HMRC, because it is a significant part of the overall tax take, and wealthy individuals have access to advice and expertise that normal taxpayers do not have. Recognising all of that, HMRC has a specific focus on that element of tax compliance, and we prepared a detailed report on how HMRC is performing in that area. UK wide, we are seeing a significant increase in the tax take from that group of taxpayers. It is a large part of our work with HMRC across the board, but our recommendations in this report are focused on the divergence in Scotland.
In HMRC’s outturn figures and the Scottish budget, the amount that is estimated for income tax and so on is based on forecast revenues by the SFC and the Office for Budget Responsibility. HMRC reports on the reconciliation and the actual figure three years later, which can be a plus or a minus. The reconciliation that has been applied to 2026-27 is plus £406 million, because tax receipts were higher than forecast.
The Scottish Government can borrow only a limited amount to smooth out the effects of possible forecast errors in income tax receipts. The amount that the Government can borrow is roughly £600 million, which rises with inflation every year. That could be seen as an arbitrary figure; I am not at all sure how it was reached. What is it based on? Given that income tax receipts are rising faster than inflation, is the limit still fit for purpose? I am not sure whether the Auditor General or witnesses from the NAO should answer that one.
I am happy to start, and others may want to come in. I am not sure that the question is for either of us; it is perhaps more for the Scottish Government, as it is about the Scottish Government’s engagement with the UK Government.
You will know that the limits were changed in the most recent update to the fiscal framework. From the Cabinet Secretary for Finance and Local Government’s correspondence to the Finance and Public Administration Committee, we know of the Scottish Government’s desire for the fiscal framework to be reviewed earlier than 2028, which is when the timetable sets that out.
You are right that the limits have changed in line with inflation. Whether they are the right limits with the right associated borrowing powers is a matter for the UK and Scottish Governments to agree on through further iteration to the fiscal framework. However, it is clear that the tax take amounts are changing, and doing so significantly. The NAO’s report sets out that the volume of tax collected in Scotland—and in the rest of the UK—is increasing. The alignment of the associated risk and borrowing powers will be a key part of the consideration of any changes to the fiscal framework.
Would you not consider it part of the audit function to comment on the adequacy or otherwise of the limit?
There are two things to consider. One is that aspects of that issue are about policy. As you know, it is not for us to comment on the merits of policy. In the previous iteration, we saw an evolution to recognising that the limits will be subject to inflation.
Equally, it is important that the powers are consistent with the level of risk; anything beyond that is tripping into the territory of asking what the components of the fiscal framework would be.
The limit is quite important. In effect, at the end of the year, we are balancing our books on that basis. I would have thought that there would be a comment in the audit report about how the limit was functioning, whether it was functioning well, whether it was adequate and whether it appeared to be doing its job.
You are right that the Scottish Government has to manage the volatility between years that comes from the fact that reconciliations come in. Depending on what the outturn results are, some of those will be positive reconciliations, such as the ones that we are talking about here, and some will be negative.
In the report, we mention that, so far, the Scottish Government has been able to stay within the borrowing powers and manage the situation within years. However, we recommend setting out a more detailed strategic response to how it will manage reconciliations together. If it knows that, over the medium term, some reconciliations will move it up and others will move it down, how will the Government use its borrowing and reserves powers together to manage that and to level out the flow between the years? We recommend that the Scottish Government should look at that in order to have a better understanding of how it will manage those reconciliations over a few years.
The details of the extent of the boundaries on borrowing and so on are decisions for the Scottish Government to negotiate with the UK Government, which it last did in 2023.
I will move on to—
Colin, I will bring you back in later if we have time. I am conscious that we are up against the clock. The deputy convener has one question that he wants to put to the NAO.
I will do so quickly, in the interests of time. Thank you for joining us, Mr Davies. I want to carry on with that theme and ask about a comment that you made to the committee in last year’s evidence session. You said:
“One of our big critiques of Government generally, including HMRC, is the surprisingly low level of investment in evaluation of the impact of different policies.”—[Official Report, Public Audit Committee, 19 February 2025; c 10.]
Is that still the case? Is it still a concern, or have you seen any improvement in the past 12 months in the level of investment that the Government or HMRC has made and in the understanding of how tax-divergent and tax-differential policies in Scotland have an impact on the Scottish budget?
We have seen a greater focus and more activity—that is, not only discussions and planning. For example, there has been a move to annual checks in some cases when there had not been annual checks until then. We recommended proceeding to complete a planned evaluation as quickly as possible, and that exercise is under way; it is an additional commitment that has been agreed to since we previously met.
I hope that the answer to your question is that increased effort is going in. There is a question about how quickly the evaluation will be completed, which is why we made our recommendation. We can use the results to work out where there is a genuine cost benefit in having more detailed information about how some of the trends are diverging in Scotland and where such an exercise is just an expensive way to collect data that does not have much practical use. That is the key judgment that the evaluation will come to. There is more focused effort.
Thank you very much. Joe FitzPatrick has some questions.
The witnesses have already covered the area that I will raise a bit, so I will tighten my questions. I want to talk about compliance risk. You have said that HMRC’s assessment is that Scotland’s compliance risk is similar to that for the UK overall, given the current levels of divergence. Is there a certain level of divergence at which the risk would change?
Obviously, we do not have any separate evidence from that of HMRC, so we cannot challenge those conclusions with a different evidence base, but sheer logic suggests that, as the level of divergence increases, some of the economic incentives will become stronger. Rather than assume that the risk will not change, that puts the onus on gathering the evidence to check whether that is the case. That is why, in previous years, we have not felt the need to make the kinds of recommendations that we have this year. We think that the level of divergence is now getting to the point where, although it is still correct to say that the risk is not changing, we need to prove that rather than assume it.
When you are doing the work to assess whether the risk has changed, do you look just at the tax base side of things or do you look, for instance, at the wider social contract? There is evidence that people are moving from the rest of the UK into Scotland because of the social contract and in spite of tax differences.
Such considerations definitely stray beyond the auditor’s remit, although there is data to inform that picture. Understanding how taxpayers are moving around the UK is really important for HMRC, partly because its system depends on correctly identifying Scotland-based taxpayers through their addresses and the allocation of the S prefix to the taxpayer record. If, for example, there was an increase in the number of people who were moving from England to Scotland, that would change the HMRC workload that is associated with keeping its records up to date.
Of course HMRC needs to understand the trends. Whether that results in a risk that is different from before is a separate question, and that is where we think the evidence base needs updating through this evaluation.
I absolutely agree with Gareth Davies’s assessment, but what caught my eye in the NAO’s report was figure 5, which shows an analysis of the proportion of income tax paid by different taxpayers. I note that the top rates of tax are paid by 20 per cent of the total number of taxpayers. I agree that it is not for auditors to talk about population movements or incentives, but that perhaps speaks importantly to some of the engagement that HMRC is having with top-rate taxpayers, given how pivotal they are to the totality of tax paid.
You talked earlier about the on-going evaluation. Did something specific trigger that evaluation? What can we expect from it?
HMRC has had plans in place for that evaluation for a couple of years, and that links back directly to the change in risk because of the different tax rates between the two countries. With regard to what is planned, the issue is how quickly the evaluation can be done, because we are at the stage where, if we stand back and look at it, the difference between Scotland and the rest of the UK is now quite well established. That is one of the things that we are really pushing on.
Graham Simpson has some questions for you.
10:00
Mr Davies mentioned the S code, which is the Scottish code. I remember that we asked you about that last year. A number of companies have not been applying the S code. Where are we now on that? Has the level improved?
I will make sure that I have the right figure for that. We think that about 45,000 errors in the application of that code are being picked up, which is a fairly similar level to previous years; the figure is going up slightly.
As always with measures of when a system is picking up errors, in one sense, it is reassuring to see that it is working. In other words, I would be suspicious if the number was very low, because I would be worried about whether the control was working properly. It is evidence that a proper check is being done and that it is identifying errors, because humans make mistakes—deliberately or accidentally. From experience of auditing other systems, we would expect to see something like that.
Is the figure pointing to a trend in any direction? I do not think so at the moment. We say in the report that, when HMRC identifies a cluster of cases with a particular employer, it engages with that employer to point that out and to understand the system that the employer has in place for correctly identifying Scottish taxpayers and updating the pay-as-you-earn system accordingly. We think that that is an effective approach.
There is no evidence that that is becoming a bigger problem or that there is a higher number of problem employers in this case. The system for identifying errors and following them up at employer level is a good one and seems to be working.
Last year, HMRC said that there were more errors on the S code by large businesses and businesses in the financial and insurance sector, which quite amused us. Is that still the case?
Peter Morland may have more detail on that.
I would need to double check, but I believe that HMRC is still focusing on that area.
If it is still the case, why is that not being sorted out?
HMRC’s strategy is to target employers and educate them about the need to put the right tax code prefix in place. A lot of this is about education to make sure that they do that, and some of that will take time.
It is certainly taking time. It is not your fault; it is HMRC’s.
Of course, you would want to see an escalation if a company, despite fair warning and constructive engagement from HMRC, was not improving its performance. HMRC has ways of ratcheting up its attention. If this continues to be a reported issue from HMRC, we could follow it up in more detail next year.
That might be worth doing.
Another issue that we have raised before is missing addresses, which still appear to be an issue. What is the scale of the problem now?
That is a common issue across the tax system. People’s lives are complicated and sometimes messy, and people can have rapid changes of address in a short space of time. Systems such as the tax system struggle to keep up with the reality of people’s lives in some cases.
A level of that kind of error is inevitable in any tax system, however well it is maintained. The point is to keep an eye on trends over time. If it is a worsening problem that requires focused attention, it should get that, but I do not think that we are signalling that in our work so far. Again, it is just a feature of a compliance system that is, in essence, working to identify errors.
The issue seems to be that HMRC cannot match addresses to postcodes and people.
When there are new developments, the records need to be updated accurately with new addresses. That is the dynamic nature of property records and taxpayer records. There are not only changes as new developments come on stream; people also move between addresses. That is why you end up with gaps in the record that need to be plugged.
I do not understand why it is still such an issue.
I will ask you about one final area. The report mentions wealthy taxpayers. How would you define a wealthy taxpayer?
I mentioned earlier that, in 2025, we prepared a separate report on HMRC’s approach to wealthy taxpayer liabilities across the UK. Our report used HMRC’s definition of wealthy taxpayers as those having income of more than £200,000 a year or assets of more than £2 million. That is clearly set out in the report, and it shows how HMRC has organised its focus on wealthy taxpayers. The definition covers a wide range of taxpayers, from the super-rich to the rich. We commented quite a lot on the impact of the definition, and there was a lot of debate in the UK Parliament as to whether it was useful. Our report sets out in detail how HMRC approaches that segment of the taxpayer population.
Do we know how many of those individuals are in Scotland and whether the number has gone up or down?
Not for the purposes of this report. That goes back to the granularity of HMRC’s records. HMRC does not approach wealthy Scottish taxpayers as a separate group; it deals with them as wealthy UK taxpayers. The compliance controls that are applied to people who meet the definition in Scotland are the same as those that are applied across the rest of the UK.
According to your Scottish income tax report, wealthy people have individual compliance managers, so they get their own person to deal with. You would think that HMRC would know where people live if they deal with them on such a personal basis.
I suspect that HMRC knows where those people live. You are talking about the top end of the range that I described. HMRC uses that system; the return on investment from working with people who pay so much tax is substantial, which is why it has approached the matter in that way. In the conclusion of our report, we note that HMRC has improved its approach to dealing with the compliance of wealthy taxpayers, which is reflected in the significantly increased tax take from that group in the past few years. It is a good news story for tax collection.
We are interested in the tax take in Scotland. Has the tax take from wealthy or very wealthy people gone up in Scotland?
The Auditor General for Scotland pointed to the substantial contribution that the top section of taxpayers makes to the overall total in Scotland, which is in our report. The controls that I have described and which we think are working better in the UK as a whole also apply to Scotland. You can take some comfort from the fact that HMRC has a better approach to wealthy taxpayer tax collection than it did a few years ago.
Before I bring in the deputy convener, I want to go back to the point about the financial and insurance sector. I would have expected that, out of all the parts of the economy, employers in the financial services sector would be on top of tax codes and tax arrangements. Earlier, Mr Davies spoke about deliberate or accidental mistakes. Would it be reasonable to infer that deliberate mistakes are being made by the financial services sector?
I have no evidence on that, I am afraid. As you would expect, I will try to stick to that answer.
Your question is one for HMRC. It may have more insight than we do at the moment as to what is going on and why that sector is posing a particular problem. I am guessing, but it could be due to the sheer size of those businesses. They could be dealing with employees all over the UK and they may be struggling to identify the difference between their Scottish and English employees. The question is, what evidence has HMRC got so far about the issue? Why is that sector a problem? As you say, it does not stand to reason, because the issue ought to be well within in its sphere of competence.
Yes, and there is also the fact that the problem is persistent. That was a speculative defence on the sector’s behalf, but we could also mount some speculative prosecutions. [Interruption.]
Apologies for the noise; it has nearly finished—it is not my stomach, although I have not had my breakfast.
Mr Davies, I appreciate that you will log off at quarter past 10, so I will start my line of questioning with you before aiming my other questions at the Auditor General.
One of the issues that came up last year—it comes up recurrently—is how complex it is for the Parliament as well as the Government to work out how much the Government can expect to receive from its tax policies versus how much it actually gets when the outturn is delivered. It is also complex to figure out whether those policies are producing, over a period of time, the levels of income that the Government thinks are required for its spending decisions.
So much of that is based on estimates and assumptions that are made by various bodies, including HMRC, the SFC and others. How do we get to a position of relying less on estimates and assumptions and instead get more accurate forecasting so that the Government has a better idea of how much its policies will raise financially?
It is always true that the amount that will be received from income tax is inherently an estimate for the year. That is because tax rates are set before the beginning of the year, and changes to the economy—economic growth going up or down, or jobs being created or lost—can affect the amount of income that people earn and, therefore, the tax that they pay.
The Government produces a budget, and it is in everybody’s interest that the budget is based on data that is as sound as possible. The key point is to use the best data that is available to set the budget, which is where the Auditor General’s work on the Scottish fiscal framework comes in—he is better placed to comment on that than I am.
Every budget process inherently involves estimates, particularly for income tax. As we point out each year, an interesting feature of the Scottish system is that, even though the vast majority of the outturn is based on the actual tax that has been paid by taxpayers, 5 per cent of the outturn figure is still estimated. That percentage is slightly reduced this year, but it is still about 5 per cent. That is because HMRC does not have the granular data that it needs to be absolutely precise about the allocation of some of the smaller elements of the total tax figure and therefore uses an estimated basis for allocating a share of the UK total to Scotland.
We think that there is scope to keep reducing the remaining estimated percentage of the outturn. Clearly, we should be able to get as accurate a figure for the outturn as possible, and we think that there is more that can be done with granular data to reduce the level of estimation.
Thank you—that is very helpful.
I will briefly revert to a previous line of questioning about identifying the volume of taxpayers in certain tax brackets. We know that, due to Scotland’s population, there is a smaller number of people who pay the top rates of income tax. However, 18 per cent of taxpayers pay more than 65 per cent of all tax, so they are an extremely important group of taxpayers. I find it odd that no one can answer the question of how many taxpayers in Scotland earn more than £100,000, £200,000, £250,000 or £500,000, given how much tax they pay. It is a relatively small group of people—you could probably fit them in this room. How do we get that sort of data? It would only take one or two of them to exit the tax system in Scotland for us to have a problem.
That is a question for HMRC, when you get a chance to question it, because its system is what allows it to identify different groups of taxpayers. You will not be surprised to hear that that kind of question also comes up a lot in the UK Parliament, and HMRC’s answer is always based on the limitations of its system in giving that kind of data.
10:15
Is that an inherent problem with the contract that HMRC has with Governments? Presumably, HMRC’s job is to collect the money and produce the data. Is it not giving us more data because it has not been asked for more? Should the Governments be asking for more?
That is a big question. Clearly, the Government as a whole, and HMRC as part of that, has a big problem with legacy data systems that were designed a long time ago and have been left behind by modern technology developments. The Government is not a greenfield site—it cannot just build an entirely new system using the best available modern system; it has to upgrade over time. That is expensive, so it has to do that in stages, as it can afford to.
There remains a significant digital transformation job for HMRC, along with many other bits of Government. That is a big focus for our work. We report regularly on progress on making text digital, for example. The long-run answer to such questions is that modern systems can make data analysis easily available. However, how do we get from where we are, with many different old systems, to a comprehensive suite of new systems that use the data more sensibly?
That is very helpful. Thank you. You will be pleased to know that I will leave you in peace now, Mr Davies.
Gareth Davies and Peter Morland, if you need to go off to your next meeting, we understand. I do not think that we as a committee have any further questions for you, so you are free to go.
Thank you very much.
Thank you very much again for your contribution this year—it is much appreciated.
Auditor General, it is over to you now, if that is okay.
I want to ask you about the statement in your report “Financial sustainability and taxes” that Holyrood’s budget has been boosted by more than £4 billion since the introduction of devolved taxes nearly 10 years ago. However, that figure is, and I quote your words,
“significantly less than the additional tax raised over the same period. This trend is set to continue.”
What do you mean by that?
That goes back to some of the questions from the convener about how the tax that is raised in Scotland does not automatically flow through in its entirety to the Scottish budget. That is what we have coined in various reports as either the economic performance gap or the tax performance gap, because of the mechanisms in the fiscal framework around the relative growth of the UK economy compared with that in Scotland. In our November report, we made the point about the need for a clearer alignment between tax strategy and economic strategy in Scotland, and for the Scottish Government to set out more clearly in its economic strategies how it intends to close that gap.
This is getting into the territory of speculation, but it probably will not be possible to close that gap altogether, given the nature of some of the structural differences that exist between the Scottish economy and jobs and those in other parts of the UK. That was best exemplified in the letter that I mentioned from the Scottish Government to the committee in April last year, which talked about the financial services industry in Scotland compared with that elsewhere in the UK. Although Edinburgh has a very successful financial services industry, as does London, the nature of and remuneration for the jobs in the different cities are quite different. That means that it is for the Scottish Government to think about how Scottish economic policy can continue to bring high-paid jobs into Scotland, as that would be the most likely driver of closing the gap in the years to come.
It is about more jobs, better-paid jobs and a higher tax base—are those the solutions to the conundrum of that percentage in the pound not going into the Scottish economy?
In essence, as you say, it is about growing the tax base with higher-paid jobs. Given the way in which the application of the fiscal framework between the two Governments operates in Scotland, that is the most likely route to closing the tax performance gap.
When we talked about this last year, the convener said that, for every £5 that is raised through additional tax, only £1 finds its way to the Scottish budget. You replied that that was correct. The figures that we have looked at today—£1.7 billion returning £600 million—would suggest that the situation has improved somewhat. For every £5, we are now looking at about £1.80, so it is heading in the right direction. However, are we ever going to get to a stage where, for every £5 raised, £5 is available to spend, or is that an impossible scenario?
This will be speculation. The figures will be based on economic models, particularly the forecasting that the Scottish Fiscal Commission does to support the Scottish Government and inform its medium-term financial forecasting, its tax strategy and, ultimately, the annual budget. I do not know whether the tax performance gap will ever be entirely eliminated, but, in our November report, we say that those strategies need to be more closely aligned and that the Government needs to set out transparently how its economic approach, initiatives and interventions will bridge the gap. That was a key part of our recommendations.
As you mentioned earlier, part of that is the Government being open and transparent about those numbers. I will give you some examples. This week, my office has been doing research on comments or statements made by the Government, including the First Minister, cabinet secretaries and other ministers, which found that the higher figure of £1.7 billion is often used by the Government as an example to justify its policy decisions. I am not asking for any comment on the policy decisions, but, when these issues are discussed, it is only ever the higher figure that is mentioned. When they are talking about how much money tax actually produces for the Government to spend, they very rarely use the lower number. Do you think that the Government ought to use both figures, for example, when talking about tax divergence?
As we mentioned earlier, both figures are there. If you look at budget documentation, you will see the figure for additional tax collected and you will also find what that means for the Scottish budget. However, we think that, in order to support transparency and effective scrutiny, the figures need to be closer to each other in the documentation. Richard Robinson might want to elaborate on his earlier point about our consideration of the 2026-27 budget documentation. We have seen some movement, but there is not yet clear positioning of the tax take relative to what that means for the Scottish budget. There are more opportunities to improve transparency on that point, but I will turn to Richard briefly if he wants to add anything.
I probably do not have too much to add to what has been said. Both figures are important. The figure for tax raised gives an indication of what the impact of the policy-only environment would be, and the net figure gives more information about the spending power consequences. The SFC will forecast those figures forward. Those are available within the SFC forecast evaluation, and they are included in the budget documents, as the Auditor General said. The closer together those two figures are presented, the easier it is for those undertaking budget scrutiny to understand what each of the figures means and how that information supports what the Scottish Government wants to do in relation to fiscal sustainability and annual budgets.
The problem relates to how the public can scrutinise those figures when ministers only ever quote one of the numbers. The example that I have is probably the worst case of it. The First Minister said that
“According to the Scottish Fiscal Commission, the policy choices that we have made will raise up to an additional £1.7 billion in this financial year”,—[Official Report, 22 May 2025; c 20.]
and that that would “help to pay for” a long list of things that he went on to say.
The general public would be listening to that and saying, “'Fine—I may not necessarily agree with tax divergence decisions made by the Government, but it raises £1.7 billion to spend on all these free things that we apparently get.” However, we know that that is not the case. Unless both numbers are quoted by senior members of Government, the public will have no idea how accurate they are or whether they are relevant to the conversation.
I go back to key message 1 in our November report, in which we referenced both those figures. The Scottish Government expects to raise £1.7 billion from Scottish income tax as a result of its policy choices, but only £616 million is projected to benefit the Scottish budget. We think that there is a need for more transparency about the difference and why it exists. I go back to the key recommendation for the Scottish Government, which is to set out more clearly how it intends to bridge the difference between those two numbers through its economic strategy in the years to come. As ever, it is our intention to follow up and continue to monitor and report on progress on how that is set out and the steps that are taken to close that gap.
My final question is on a comparison with last year and the meeting at which you talked about the issue. We get a feeling that things are improving in some areas—I hope that they are—but if we look at statistics from last year that are relevant to what you have just said, we see that Scottish average earnings last year grew 3.1 per cent less than those in the rest of the UK, and employment rates grew 3.2 per cent less than those in other parts of the UK. That was 12 months ago. Do we have any idea whether we are starting to catch up in those areas?
I do not have that detail to hand—I do not think that either of us does. We can either look at that or signpost the committee to the most up-to-date figures.
Included in that would be economic inactivity, which is another key area where there is some divergence, particularly among younger people. There is quite a substantial difference in inactivity levels. Those are the sort of statistics that we would want in order to give us a feel for whether there is improvement—whether the gap is narrowing in any way or getting worse. Both of those areas are relevant and important to us.
We were short of time earlier, but do any members of the committee have further questions that they want to put now? Colin Beattie is indicating that he does. With a degree of brevity, Colin, over to you.
If you recall, Auditor General, the committee has, in the past, challenged some of the adjustments and estimates, and has looked at the impact and the overall figure in the aggregate. We had a useful private briefing with HMRC last year, which was really helpful in giving us some background. I notice that, this time, HMRC has greatly reduced its references to adjustments and estimates. It no longer details them, but instead provides an overview. The NAO report details some of the estimates and adjustments, but by no means at the same level as before. Do you have any more detail on how those estimates are audited? Who looks at them and checks the logic in them? On what basis is there an assurance that the estimates are accurate?
This will probably take me into territory of commenting on the specifics of what the NAO has done, when the NAO is far better placed to describe the steps that it has taken. If I can find the right reference, I will refer to a table in the NAO report that sets out some of the estimates and deductions—it is figure 2, if I am picking the right table. There are a couple of points, one of which is relevant to your question. The process will be undertaken as part of this audit, and also in the NAO’s overall audit of HMRC. I hope that it is helpful to the committee to know that our work is not to reperform the NAO’s audit, but, rather, to make an overall assessment of the adequacy of the approach and whether the NAO’s conclusions are reasonably based. That is what we have done. There is perhaps a limit to the extent to which either of us is able to give you the detailed response that you are looking for. For good reason, I am reluctant to represent the NAO’s work.
10:30
As the Auditor General indicated, the Scottish outturn figure in figure 2 is largely based on actuals as opposed to estimates. We can see £115 million under the estimates column in figure 2, with £17 billion as the total figure. There is some assurance in the way in which the estimates are collected, which is after the self-assessments and so on are completed.
There is also a slightly more technical point, which is around how these estimates are proportioned. As the NAO’s C and AG was saying, because there is a similar strategic picture of risk, when these estimates are applied, they will be applied in equal proportions to both. On one level, that does not matter so much, but, when we are talking about how that works through the fiscal framework, it makes that more marginal, because it is applied equally or proportionally across Scotland and the rest of the UK.
Okay, I will leave it at that. I think that my other questions are for the NAO.
We can follow up on any outstanding questions in correspondence with the NAO.
My final question goes to the point about the cost of administering the Scottish income tax and the service-level agreement that exists. I think that there are committees in place that monitor that and look at the quality of data. When we consider these reports each year, it always strikes me that the administration cost is quite low—I think that it is £1.2 million. I always start to wonder whether, if a higher administration fee were paid, we could get more Scottish-level data out of the process. I do not know whether you have any observations on that, Auditor General.
I recognise that the scale is quite stark with regard to what is paid, relative to the amounts that are received. It is probably more a question for the NAO and the Scottish Government about whether they are getting value for money.
Further options are available to the Scottish Government in considering whether it wishes more activity to be undertaken and further analysis to be done on the Scottish-specific tax gap. The figure of £5 million is quoted as the annual charge that HMRC would levy on the Scottish Government to do that. It is clearly important for the Scottish Government to think about what comes next in supporting the accuracy of estimates and forecasts, and, ultimately, the overall amounts that are collected. We certainly note the figures, and you can see from part 3 of the NAO’s report that it considers that the figures are fair and accurate.
The Scottish tax gap is the difference between what should be paid and what is actually paid in tax. There have been discussions over the years about whether HMRC would start to collect that Scottish-level tax gap data. There is a suggestion in this year’s NAO report that it is considering that. Do you get any sense of how close we are to that figure being calculated and being part of the annual report? This committee would find that very useful indeed.
It depends on the next steps that the Scottish Government chooses to take and whether it commissions HMRC to undertake that work. As the committee heard this morning from the C and AG, given the divergence between the tax regimes in Scotland and those in other parts of the UK, and particularly the evidence that you heard about the opportunities that the wealthiest taxpayers can take up, there is a level of understanding about the tax gap to support assurance that the money that is due to Scotland is being collected, along with the other compliance arrangements. You can see from the C and AG’s recommendation that, in pace and totality of focus, that looks like it is soundly based.
Thank you very much. It has been a useful, albeit brief, session this morning. There are further elements of the NAO report that we will want to follow up. For now, I thank the Auditor General and Richard Robinson for their evidence this morning.
I now move the meeting into private session.
10:34
Meeting continued in private until 11:33.