 Good morning, and welcome to the eighth meeting of the Public Audit and Post Legislative Scrutiny Committee in 2017. Can I ask everyone to switch off their electronic devices or switch them to silent mode so that they do not affect the committee's work this morning? Can I ask everyone, particularly this morning, to double check that they have done that? Before I begin today's meeting, I am sure that all members will wish to join me in sending our condolences to the families and friends of those who lost their lives in London yesterday and in paying our respect to all those who displayed outstanding bravery in trying to help others. There is going to be a one minute silence at 9.33 this morning across the Scottish Parliament in solidarity with our colleagues at Westminster, and at 9.33 I will invite all witnesses, members of the public in the gallery and staff to join committee members in observing the silence. We will then resume our business. I would like to tell everyone present that the cameras will be on us during this minute silence and to bear that in mind as you observe that. I am just going to wait until 9.33 and then I will invite you to observe the minute silence with me. Can everyone please stand? Thank you. Item number 1 today. Our first item is to decide whether to take item 3 in private. Do members agree? Item number 2, we will now take oral evidence on the Comptroller and Auditor General's report entitled The Administration of the Scottish Rate of Income Tax 2015-16. I welcome to the meeting Stephen Corbishley, director at the National Audit Office and Caroline Gardner, Auditor General for Scotland. Unfortunately, Sir Amius Morse, the Comptroller and Auditor General, is not able to attend today's meeting. He was caught up in the lockdown at Westminster following yesterday's events and could not leave until very late in the evening. Stephen Corbishley will therefore provide an opening statement instead. Before I invite opening statements, I will briefly explain why we are taking evidence from both the National Audit Office and the Auditor General today. From April 2016, the Scottish Parliament has set a Scottish rate of income tax. The accounts of HMRC, who are responsible for collecting and administering this writ, are audited by the National Audit Office on behalf of the Comptroller and Auditor General. To ensure that the HMRC's collection and administration of the Scottish rate of income tax is subject to scrutiny by the Scottish Parliament, the Comptroller and Auditor General is required to report to us. In turn, the Auditor General for Scotland provides additional assurance on this audit work by also submitting a report to the Scottish Parliament. I now invite an opening statement from Caroline Gardner to be followed by an opening statement from Stephen Corbishley. As you know, the Scottish Parliament's financial powers are changing substantially with new responsibilities for taxes, social security and borrowing through the 2012 and 2016 Scotland Acts. My latest report on progress across the new financial powers is published today and I look forward to briefing the committee on that at a later date. The purpose of today's session is to look at the auditing of one element of the new powers, that is the Scottish rate of income tax. Specifically, the reports that the committee has before it relate to 2015-16, during which preparations for implementing the first of devolved income tax powers continued prior to their introduction in April 2016. In considering those reports, it is important to bear in mind the respective responsibilities of those involved. Firstly, for income tax itself, HMRC collects and administers income tax and is responsible for the project to implement the Scottish rate. The Scottish Government is responsible for funding the project and for seeking assurances that the new system collects the correct amount of tax. Secondly, the respective responsibilities for auditing. As you said, convener, the National Audit Office audits HMRC's accounts and the Comptroller and Auditor General is responsible for reporting to the Scottish Parliament on HMRC's administration of Scottish income tax. I report to the committee to provide it with additional assurance on the NAO's audit work in line with a recommendation from the predecessor public audit committee. This is the second year of this arrangement. In summary, my report says that I am satisfied that NAO's audit approach was sufficient and robust and covered the key audit risks. I am also satisfied that the findings and conclusions in the Comptroller and Auditor General's report are reasonably based. I will now pass on to Stephen Corbishley on behalf of the Comptroller and Auditor General to provide opening remarks on the outcome of the audit work at HMRC on the administration of the Scottish rate of income tax. Convener, good morning, everyone. Can I just start by passing on Sir Amos's apologies for not being with you today? He did want to be here but, obviously, circumstances, which I am sure you understand, meant that he was unable to to make it today. I am deputed to take our report forward with you today, so please bear with me as I take us through this. As you have already pointed to, the CNG's role in respect of HMRC is to scrutinise both the accounts that are produced by HMRC and also our value for money programme within HMRC and the report to the UK Parliament on those activities. However, as I have already alluded to, because of the devolved nature of certain activities within HMRC, particularly in relation to the Scottish rate of income tax. The amended Scotland Act 1998, the CNG, has powers to report to both Westminster and the Scottish Parliament in terms of HMRC's administration of the Scottish rate of income tax, which is why we are here today. This report is our second report, where we are considering the implementation phases of HMRC's administration. Our first report was with your predecessor committee covering the 1415 activity. In today's report, it looks largely at the 1516 activity, which is the last formal year of the implementation process from HMRC's perspective of the implementation of the Scottish rate. Our key findings in the report, which we will talk you through, does look at the progress, in particular to some of the key risks that HMRC has faced in implementing this project. 1617, which I realise is the year that we are just about to complete, is obviously the first year where there are devolved powers. We will look at the effectiveness of the controls and the implementation as we audit the 2016-17 accounts produced by HMRC and readiness for our next report later on. I may be able to talk you through, but some things are still to be resolved with HMRC. I am happy to take any questions, of course, on our report and to put our report into context, although, as you probably appreciate, some of the answers may be better directed to HMRC in due course. Finally, just to support the Auditor General, we have worked very closely with her and her colleagues under our memorandum of understanding in order to help the Auditor General gather and provide the assurances over the work that we have done as well. It is safe to say that it is a very good, challenging and productive work in relation to which we value greatly in terms of putting the report together. I am looking at page 11 of the report in connection with the error in the design of HMRC's taxpayer identification. HMRC does not have a great track record on IT. Was there any cost to the Scottish Government of this error? Yes, there was. HMRC told me that they have not passed that on to the Scottish Government, and the cost was around £150,000. £150,000? Yes, in putting it right. The Scottish Government did not have to cough that up, so to speak. HMRC tells me that that is the case. Part of that seems to indicate that, on page 16, on 2.5, you are talking about approximately 850,000 records where the Scottish postcode was not flagged as Scottish. And then, just below it, it says that an estimated 420,000 of that 850,000 were potential Scottish income taxpayers. Why is there that difference? You identify 850,000 who apparently could be Scottish taxpayers, and then suddenly that halves. HMRC holds records. HMRC is, of course, a tax-gathering authority, but it is also a benefit, a payer of benefits for tax credits and child benefit as well. It has records across the whole suite of its activities on people who might be regarded as Scottish taxpayers. Then that distillation, that further data analysis, would hone that down to the £400,000 figure that we have talked about here, who are potentially Scottish taxpayers as opposed to having a record in HMRC's database. So, are we saying that the balance there, the 430,000, are actually people who are on benefit as opposed to being potential taxpayers? Or they will have some other record, which doesn't necessarily mean that it doesn't lead them to being potentially Scottish taxpayers. Yes, in a nutshell. So, if they are on benefit and get a job, will HMRC pick that up? That is what this process was designed to look at. So, when the database has been properly cleansed, HMRC would argue that that enables them to be able to maintain that database, that would pick up their basis that they are Scottish taxpayer. I think that that is more of a question for HMRC in terms of that flow of when people move from benefit into being a taxpayer. It is obviously not a linear process. I would say that it all seems wee bit hit and miss as if everybody is feeling their way, because I am looking again at page 18, 2.12. We have seen that we are very reliant on employers doing the right thing in order to maintain the list of taxpayers. A lot of this does depend, pay as you earn, of course, is managed by employers on behalf of HMRC through the real-time information system. So, you do rely on, awful lot, the tax system does rely for that particular tax stream on employers doing the right thing in terms of its flow of information to HMRC. HMRC clearly has a duty to ensure that they are capturing the people in the right way to flag the Scottish taxpayer. However, it is not all about HMRC. There is a lot of discussion with the employer community, so it is making sure that their information is right. When HMRC comes to cleanse their information, they rely a lot on employer information to help in that cleansing process. Again, on page 24, 2.31, you are saying that 75 per cent of large businesses have a workforce that is partially Scottish. Originally, there was some debate at the very start of that. How big businesses operate across the border would actually be able to identify and flag Scottish taxpayers? Has that been resolved? I think that we can say that that is work in progress from HMRC's perspective. A lot of part of the reaction to the issue on the taxpayer identification exercise telling last year was actually how they can use, for example, test their data, HMRC test its data alongside information from, for example, the larger employers such as the Scottish Government, for example. Again, it is using that data set and comparing that data set with information that is held on HMRC's own records. They will continually refine the information that they would use to identify Scottish tax base. It is not just a one-stop, it is not just a big cliff. Clearly, that database needs to be maintained day on day, year after year. There is an on-going maintenance of the database. Given all those elements, how confident are we that we have captured a real number for the income tax that is being collected in Scotland? I think that some place in here there is something like a 4 per cent error rate looked at, which is indicated. Again, in paragraph 2.31, you are saying that 25 per cent of employers there seems to be an issue with the responsibility to log workforce changes. How close are we to getting a real figure on this? The easy answer for me is to say that that is what we will look at during 1617. HMRC has told us that they are continually refreshing, continually looking again at their database to ensure that it achieves the clear objectives that you rightly set out for this particular project. If you are asking me to give you assurances at this point in time, I will defer that till we have done some hard audit on the figures as they start to come through based on the 1617. They are moving in the right direction. That is probably as far as I would go at this point in time, but there is still some work that they need to do for that on-going maintenance of their database. You are doing the audit of all this. Do you have a feel as to what the possible real error rate is? No, we do not. I think that what shook us all in all of this was that the 420,000 misidentification, if I can put it that way, out of the first scanning exercise that happened a year ago, seemed to be just a procedural formulaic issue that led to a number of people being missed off the initial mailing lists. HMRC's action will be to bring that down to a reasonable figure. You will never get, I guess, 100 per cent accuracy in this part of HMRC's activity. That is just the nature of the flows of people moving in and out, moving house, moving in and out of Scotland or from Scotland into England or England to Scotland and vice versa. You will never get 100 per cent accuracy, but you want to get it to a point that is materially correct. As an auditor, what would you say a reasonable percentage accuracy would be? You are testing me a little bit on that. I would not like to be pressed on that at this point. I will say that I would not like to be pressed on that at this point. Ross Thomson, thank you very much convener. I was wanting to pick up on one of the key findings around the future divergence of tax rates in thresholds between Scotland and the rest of the UK, which is on pages 23 to 24. It states that there could be a significant amount of tax at risk. Could you explain to me what you mean by a significant risk? A significant amount of tax at risk? I think that what we are saying here in the scope of the risk is that, as the rates and as the thresholds diverge, which inevitably—we have already seen that in 1718—it is absolutely essential that HMOC maintains that database of its Scottish taxpayer records or else down the line, when it comes to the block grant calculations and so on, there will be a miscalculation and a lot to catch up on as HMOC regularises or reconciles the process down the line. It all depends on the records that you have and how those are then extracted to do those calculations for block grant purposes. That is the risk that we are talking about. That is the risk that HMOC understands. They know that they have got to get down to a reasonable level, how reasonable we have just discussed it, of course. The report highlights that there is a significant amount of tax that could be at risk. In looking to the future, there are potentially two other risks. If you look at either political decisions to increase the rate of income tax or if you look at the UK Government, who are saying that the UK threshold will rise in line with inflation, you could look at Scotland's potentially paying £1,000 more in tax over the next five years. Do you know what work HMRC is doing to mitigate against that risk? Well, those are political decisions to vary tax, particularly when you get a national divergent that there is a risk there. HMRC may argue—I do not want to sound as a mouthpiece for HMRC, but HMRC may argue that keeping track of your taxpayer records, keeping track of who is paying tax in the right way is fundamental to following through the financial consequences of those political decisions. How confident are you that they have the ability to do it? I think that we have seen quite a lot of work that has been done since the original lack of identification a year or so ago, and continually refining, getting other data sources, checking those data sources against their own records. It is an on-going maintenance issue for them, and it is not a static position by any means. Providing that they have the on-going process in place and that it works to maintain the accuracy of their database, that will address the issues that you have talked about, because, as rates diverge, it becomes much more important to ensure that they do that. I think that HMRC would argue that this is an on-going discussion that they are having with the Scottish Government in terms of how they carry on refining that within the cost envelope that they have got. Will you be doing to monitor how they meet those challenges? Are you going to continue to do that? Yes, we will. We have an on-going. Our statute requires us to report to both the Westminster and the Scottish Parliament in terms of that on-going development and on-going debate, which we will do annually, of course. Although it is slightly different, do you know if there is any work being done in relation to the wider economic impact of those risks? If we are, if there is a significant risk roundabout tags, do you know if there is any work being done to look at the wider economic impact of that? I would probably pass on that myself. I think that that would be a matter for you to take with the thought of the Scottish Government. I am sure that the answer is yes, but I would not have the evidence to stress that. A very quick follow-up to Ross's question, just to help me out. We have a slightly divergent tax system already. It is not inconceivable that that will continue to diverge. What I am reading is that this is based on addresses. Where I live determines where I pay tax. Is it not open to me, if I live within commuting distance of the border, to relocate to a lower tax jurisdiction if I believe it financially beneficial? I am not here to act as your tax adviser, but yes, that could be a legitimate response if you felt that to be the case. If HMRC would argue—I think that we have talked about this a little bit before—the residency test is not necessarily just about an address, it is where your home is. If your home, if your families are based on the English part of that border and you may work in Scotland but you do not necessarily live in Scotland, you would be a classed as a rest of the UK or an English taxpayer for the purposes of this exercise. Where HMRC can prove, or you are shown to be, that resident test is that you do live your home is in Scotland—that is where your children go to school, for example—that is where your family live, then one of the tests that HMRC would apply in those cross-border issues would be a whole series of factors, such as where your family live. At the end of the day, if it comes down to numbers of days, it will exercise that formula as well, but it is not just a simple matter of just number of days, it is where you are deemed really to live. There is an issue there, of course, of whether it is avoidance or whether you are trying to do something that defraud the tax system. I am interested in the—when this gets set up, there are a significant number of Scottish taxpayers who are missed, as Colin Beattie pointed out. That was then notified to HMRC by other stakeholders. I am slightly troubled by that because wouldn't you have expected that when such a system was designed, HMRC would have put in some kind of check and balance that said, it is no working, no, let us check it is working. Talk to me about that. I agree. I could not really say any more than that. The cause of this was when they did their original interrogation of the systems, the parameters that they originally used to go through their whole, this is what we talked about earlier on, their whole record dataset did not—it missed out an aspect of the information that they assumed the postcode field was filled in, and therefore it would take the information based on that field. Not all information was input with a postcode involved, and that is why they missed. We have pointed to—we have done a little diagram in figure four in our report—of, dare I say, a back of an envelope assessment, that if they have taken those factors into account, they may have judged that, in reality, the actual number of tax places that they hit was far below their tolerance level, and therefore that should have raised a long bell in their heads. They reacted pretty quickly following that stakeholder conversations in December and January and realised that they had to fix that. I guess you, as well as other colleagues, did that straw poll in an office saying, have you had your letter yet? That was part of the debate that they were recipient of as well, and then they recast their formula to fix the problem. It is not extremely concerning that it is used the word they assumed. They should have raised alarm bells. These are vast sums of money that we are talking about. This is the revenue that the state has to spend. Just moving on from that, when they found out that they built an interim solution, why was not the interim solution already built just in case? I think that this is part of HMRC's IT systems, our lodgy legacy-based systems. Obviously, HMRC has spent an awful lot across the whole aspect of its tax base in refining and augmenting the systems that they have. There is a schedule of developmental work that needs to be done, of which this will be part of. It is not just a matter of pressing a few buttons and making it right. There is almost a profile of IT work that they need to have in place of which this will be part to correct that. That is why they had to put an interim solution in place and prior to the final solution here. Of course, we will look at that strategic solution that came into effect in October. It has taken a long time to get to that point because of all the other work that they would have to do in this space. Finally, and I may seek to come back in later, if you do not mind, convener, on all that you have said, given the sums involved and the importance of what we are talking about, this is funding what we do all day, can HMRC guarantee that this is not going to happen again? Well, my glib-cret answer is, I think that is a matter for you to put to HMRC. HMRC have told us, though, and we will test this over the next few months, that I have talked about already how they refine their database of information. They have picked up on that missing 400,000 people that they did not identify initially and are now continually revising that database of information and are almost getting to a funnel. If there is an acceptable number of people here, you will never really capture all sorts of other dynamics. However, it is how successful they are in all those refinements and that on-going maintenance of their database. HMRC have told us that we have done an awful lot of work since then and learning the lessons from all that. Checking our database, our records with information held by the significant employers in Scotland, so we are continually checking that someone that we know is working in the Scottish Government does have a Scottish flag on HMRC's database, for example, that will pick up the tax implications. If that continual data cleansing will make that right, all that the success of the project will rest on, I should say. Monica Lennon, thank you, convener. Good morning. Just to go back to the 420,000 taxpayers who did not receive the initial letter in December 2015 that 2.45 million individuals did receive, that potentially creates a less informed group of taxpayers. I wonder whether you are able to say what are the practical implications of that. Forgive me for the practical implications. For the 420,000 people who did not receive the letter in December 2015, I appreciate that they received a P2 coding note at a later date, but what are the issues for people in that situation? Part of the fix is part of HMRC's. It has now been identified as a Scottish taxpayer. The rates that they will be subject to will kick in automatically when it comes to tax calculations for those individuals. Now that they have been captured, they fall into the normal tax assessment processes. What we have pointed to on page 19 of our report in figure 6 is what are the purposes of those letters in the first place. It is clear that each individual taxpayer, as we all are, has a responsibility to keep our records with our respective tax authorities up to date. That is the first thing. There is a duty on us all to make sure that information is held. What we have pointed to in this diagram is that you can see in the red bars when the letters were issued. However, if you look at the yellow chart afterwards, the spikes in number of hits that people are updating on HMRC's database, their records, did not seem to necessarily follow when they got a letter informing them that they were a Scottish taxpayer. You can see in there that there are a number of hits following way after receiving letters. There is an on-going communication campaign that HMRC has and will have in place through other sources, not just through issuing a letter that is part of their campaign to make sure that people are doing the right thing. That is the same across the whole of the UK, of course, and making sure that that information is kept up to date. Practically speaking, it is good that they have told people the consequences of what would happen and the importance that they attach to individuals keeping their records or making sure that their records are up to date with the tax authority. However, it is not the only means by which HMRC has communicated the impact of this, or will continue to communicate the impact of this. So, can we be quite confident, as a committee, that no one has been disadvantaged as a result of being less informed than other people? We have pointed to the fact that a number of people and those people did not get the same information. The coding letters and the process from that interim solution did not get the same information that the other £4 million might be at a disadvantage in not getting those letters. On page 6 of the report, it has key facts. One of those is that £8.4 million is the amount reimbursed to HMRC by the Scottish Government for implementing the Scottish rate of income tax. Is that correct? In this in the 15-16 financial year. Okay. So, the Scottish Government has paid £8.4 million for the implementation of strut. Is that correct? Correct. Okay. I am looking at the cost of that as compared to what we raise in tax, because the Government here in Holyrood have set tax rates at the same as the British Government, with the exception of the threshold on the 40-pence rate. Can either of the witnesses tell me how much is going to be raised in the next financial year by the Scottish Government not raising that 40-pence rate threshold? I cannot. We can tell you. I do not have the figure before me at this moment, convener, so I will come back to you after the meeting. Okay. I know that you are not in the business general of giving rough estimates, but would it be more than £8.4 million? Absolutely. The estimate for what the Scottish rate of income tax would raise in 2015-16—the reconciliation has not yet been done—was £4.9 billion. It is important to note that the preparations that are put in place for the Scottish rate of income tax will be the same arrangements that will support the full devolution of non-savings, non-devidant income tax from this April, when the amount involved is forecast to rise significantly to more than £10 billion. You have that sort of scale of comparison. Correct me if I have not understood correctly, but that £10 billion is the whole of the Scottish rate of income tax. I was looking for a rough estimate on the figure that will be raised by not raising the threshold. It seems to me that tax payers in England will be paying the threshold at £45,000. We will be paying it at £45,000 here in Scotland, and tax payers in England will pay the £40,000 pension rate at £43,400. It is quite a small divergence. How much would be raised from the failure—well, not the failure, but not raising the threshold here in Scotland? We have a figure. I do not want to mislead the committee by taking a stab at it here. I will come back to you very shortly after the meeting with what the estimate is. Would it likely be more than £8.4 million? My recollection is that it is larger than that, but seriously, convener, I think that it would be more sensible to give you an accurate figure than to take a stab at it. You can probably understand us to general. The thing that I am getting at is that have the decisions that the Scottish Government has taken on tax this year, do they raise the budget in any way? I understand that the cost of implementing strut and tax income are probably in different budget lines, but does the cost of implementing this is just offset by what the Government has done on that threshold? That is the reason for the answer that I gave you initially. The arrangements that have been put in place will cover the collection of most income tax in Scotland from April this year. There will be decisions in future, I guess, about rates and bands up and down the scale, which will have an impact, but I am not sure that it is particularly relevant to pick out that one shift in the threshold for 17, 18 in comparison to the £8.5 million. Again, Steve is looking to add to that. I was just going to build on that. To be clear, the costs that we are talking about here, which over the period of not just one year but over the period of time could be approximating around £30 million in implementing the Scottish rate of income, is not just about the marginal changes in thresholds? No, indeed. For example, in figure 1, we have pointed to that 10 per cent as well, which is the bigger figure. That HMRC, a commission to the £4.9 billion that Caroline has already talked about, is what would feed into that cost calculation. Can I ask—maybe this is a question for Mr Corbishley—do you know where most of this work is being done? HMRC has centres right round the UK and has several centres here in Scotland. Do you know where most of this work is being done? A lot of this work—I might be corrected in a moment, but a lot of this work is larger than it is in Newcastle office, because that is where the payroll systems are and are monitored through. In England, I guess, that is the easy answer to the question. You will also be aware, Mr Corbishley, that there has been a proposal by HMRC to close several of its centres across Scotland. My question is from your point of view, and I realise that this is a question for HMRC as well. From your point of view, are we losing essential skills here in Scotland by the closures that are going to be needed as the tax powers come in? You are right. I think that this is more of a question for HMRC. HMRC might argue that, as it moves to a more regional-based tax office structure, it has been able to maximise and improve on the skills that people have to deal with the taxpayer community. We will, in due course, assess how effective they are in that assertion. We published a report in January 2017, in which we looked at HMRC's plans for the regional office structure, not just in which, of course, Scotland plays its part, but that was looking at the high-level costing and analysis and the impact on that skill base that you have just reported. Can you give us a brief conclusion of that report? What was the national audit office's take on HMRC's move to regionalisation and closures? What we pointed to in particular is that there are savings that can be realised, and it would be important for HMRC to ensure that they secure those savings, but savings are not at the cost of the effectiveness of the service that they provide to all taxpayers, whether they be individuals or, indeed, across the whole taxpayer base. We did not focus particularly on any one nation or any particular group of regions. We looked conceptually at the whole piece, and I think that keeping a maintenance on that cost base, making sure that they can deliver the savings that are part of their business case, but also ensuring that the service and the skills that people have to hand that people are supported in ensuring that they can deliver the services that they are expected on. You say that it is the balance between getting value for money and keeping the skills basis at the national audit's office opinion that they have struck that balance? I think that that is yet to be determined from our perspective, as we see the strategy being developed. Okay, thank you. Liam Kerr. Thank you. A couple of quick things. By the error, you suggest in the report that HMRC may have created a less informed group of taxpayers. Can you just explain to me what does that mean? Okay, so if you write to four million taxpayers saying that you are a Scottish taxpayer, what this means is that you might be paying a different rate or a different threshold in due course by a small population who do not have that same information to hand in order to make their own personal decisions. As I have already mentioned, the impact of those letters may not yield that risk, but there is a particular risk there. A risk of what are the practical implications of creating a less informed group of taxpayers? The need to make sure that I, as a taxpayer, am telling HMRC of my change in circumstances, telling him that if I move addresses within Scotland, for example, I have changed my address and I need to tell the tax authority that I have changed my address. I have just moved into Scotland and now I need to tell them that I am now a Scottish taxpayer or I live in Scotland against the definitions that we have just talked about. That is the practical circumstances there. Is that helping HMRC to keep that ongoing maintenance of its data back? I see. I would like to move on to the fiscal framework. You will appreciate it, I have no accountants, so this might take some time. From April 2017, the Scottish Fiscal Commission has a responsibility to project ahead the income tax revenues for Scotland. Is there a risk that it will be unable—going back to our earlier conversation—is there a risk that it will say that we cannot accurately predict our tax base and therefore our tax take, and thus it will not be able to do its job properly? I will stumble through an answer on this one, and Caroline May will help me here as well. Any fiscal projection, you need a good basis of data in order to do that, which is your point. That is no different for the Scottish Frame, because it would be for any Treasury projections governed by the Office for Budget Responsibility. They would need the veracity of that data in order to project, so there is a bit of a risk in there, as you point, to ensuring that that is the case. I would say in my answer that, as the block grant calculations, which is part of all of this, they will continue to be refined and adjusted over a two-year period, which is part of the settlement that you would have. As new information on the taxpayer database that I have just talked about becomes available, it is not just a one-year estimate, it will then be refined for the following year, taking into account new information about how much taxpayers have paid in this regard to adjust the previous block grant adjustment. I will hand over to Caroline May. You are taking us quite understandably into the territory of the report that I have published today on the next stage of implanting the new financial powers. I will give you a brief answer, and if I may suggest that we come back to the detail a bit later on. As you say, from April this year, the Scottish Government will have control of all non-savings, non-dividend income tax, and we have seen the prelude to that in the agreement of the 2017-18 budget in this place over the last few weeks and months. Roles and responsibilities are shifting as those new powers come into place. It is the case still that the Office for Budget Responsibility is making the forecasts of the likely proceeds from Scottish income tax at this stage. The Scottish Fiscal Commission is gearing up as it becomes a statutory body to take on responsibility for producing a wider range of forecasts, including Scottish GDP forecasts, and there will be an estimate of the proceeds for income tax in there as well. Under the fiscal framework, it is the OBR estimates that will be the basis for the block grant adjustment estimates that come through, and there will then be a reconciliation 18 months or so after the end of the financial year to look at the actual proceeds against the forecast and make the reconciliation to the block grant that is needed. It is very complex, and relationships, the roles and responsibilities that involve the OBR, as well as the Scottish Government and the Fiscal Commission as it develops. I was going to ask about the reconciliation. Forgive me, Caroline, if this is something that we should do another time. However, if the reconciliation of the projected receipts and the adjustment to the block grant happens 18 months hence, what if HMRC says that there are X number of taxpayers in Scotland? The Scottish Government presumably spends or budgets against a particular take, and the anticipated block grant is set against that level. Presumably, if HMRC gets that wrong and says that there are more taxpayers, there is more tax revenue coming in than actually happens, then over 18 months the Scottish Government has less money from the block grant and less money from its tax base. It will be 18 months before there is some kind of reconciliation, at least, to make that right. Is that correct? I think that you are absolutely right that the identification of Scottish taxpayers is at the heart of all of this, not just the Scottish rate of income tax that we are reporting on today, but the devolution of income tax as it comes into full effect next month. It is why this report is so important to you. I think that the piece of assurance that we all have is that 2015-16 was the year before the 10 pence Scottish rate of income tax came into effect and the problem was identified at that stage. I think that that has been a useful wake-up call to HMRC and to the Scottish Government about the importance of getting this right, and you have heard from Steve Corbishly about the action that they have taken to correct that. There is the continuing concern that all of us as taxpayers need to understand our responsibility to keep our tax records up to date. That now matters much more for Scottish taxpayers than it does for taxpayers across the rest of the UK. Employers need to understand their responsibilities where they have Scottish employees as a whole or part of their workforce to keep their real-time information up to date. All of that that we are reporting on now is happening at least 12 months before full devolution of income tax and any divergence between the rates and bands that would have an impact on the amount that is collected for Scotland in that direct way. However, it has been a very useful learning process for everybody to make sure that everything possible is being done to get the initial database right and to make sure that arrangements are in place for keeping it up to date thereafter. There is a whole separate set of complications about the forecast and the reconciliation of the block grant adjustments, which will come into play from the new financial year, the new tax year, and that we are reporting on today. There is also a big part of the deliberations of the budget process review group that the Finance Committee particularly is interested in. However, it is an indication of how complex that is and how important it is that the Scottish Government, the UK Government and HMRC get it right. Alex Neil Can I ask a broader question, which might be more appropriate for the HMRC, but I would have thought that it impacts on the work of auditors? We are often told that, particularly for higher-rate taxpayers, the more you increase the rate of taxation or the point at which higher rates are paid, the loss of talent leads to those people moving abroad or, in this case, perhaps south of the border. I am told, for example, that if you look at the upper rate at the moment, which comes as a kick-in at £150,000, that three out of seven of the taxpayers at that rate are actually working in the public sector in Scotland. That is about 43 per cent. My first question is, from an auditor point of view, is there evidence that the higher you charge, the higher you increase the levels of income tax—and I am talking about effective levels, whether it is through freezing the allowances or increasing the rate—is there any evidence that if you increase the effective rate of income tax, that leads to people effectively leaving the country? A lot of the public sector will be doctors who could get good jobs elsewhere without any difficulty in either the UK or abroad. I am not trying to argue that the disincentive—the alleged disincentive is only confined to the private sector—could also operate in the public sector. Is there any evidence that effective higher rates of tax lead to exporting of talent, if I can put it that way? Secondly, if there is, is there any estimate of the elasticity? At what point do people decide that it is not worth the candle? Part of the debate in Scotland at the moment is that it is okay that you are paying a bit more tax than people south of the border, but if your residents are south of the border, your children will need to pay the full tuition fee if they come to Scottish University. If you live south of the border, you will not get free prescriptions in Scotland, etc. The net impact is positive if you look at the pros and the cons, because obviously the higher revenue funds those services. However, what I am very specifically interested in is whether there is evidence that effective higher rates impact on the export of talent abroad. I will have a first fashion and then I will ask Steve to follow up. I think that there is an awful lot in that question, as I know you know, starting with the famous Laffa curve and economic theory, and that is contested. If we look down a level at what we know in more practical terms about Scotland as opposed to the macroeconomic debates about it, we know that the shape of the taxpayer population in Scotland is different from that of the UK as a whole, particularly because of the south-east of England and London. We have fewer higher rate and additional rate taxpayers in Scotland than we do in England, and it is fair to say that we do not know enough yet about the behaviour of those people, either in terms of people's ability to genuinely demonstrate that they are not Scottish taxpayers if they have a foot in both countries, as many higher rate taxpayers do, or their willingness and likelihood of genuinely making a move. As you say, doctors and other professionals often are very able to do that, given the demand for their skills elsewhere, and I am guessing the pressures on public services if we do reduce EU migration with the result of the EU referendum. We do not know that. I know that it is one of the things that both the OBR and particularly the Scottish Fiscal Commission are thinking about as they start to develop their ability to forecast and to refine their forecasts of the figures that come from Scottish income tax in future. None of this has mattered before. It has been a result of the block grant that it really has not mattered how much tax is raised in Scotland and what the behaviour of Scottish taxpayers looks like. It now becomes very important as part of the overall relative performance of the Scottish economy compared to the UK economy. Some of the data sources are available, some of them will not be. I think that when the committee has taken evidence on my report on the next stage of income tax, it may want to consider taking evidence from the Fiscal Commission and the Government and others potentially about what work they are doing to get a better understanding of those dynamics, which will be very important to making good and effective decisions about taxation in Scotland for the first time. The reverse side of the coin is the assertion that if you reduce tax rates, you increase the number of taxpayers and increase the revenue as well. If you are the Minister of Finance in Scotland, those would be some of the initial questions that I would be asking because it would inform very much what my strategy would be. If there is evidence that you lower the rates and increase your revenue, as opposed to if you have higher rates and end up with fewer people paying tax because they have moved down south to avoid it, clearly that would inform your decision about what your strategy would be. I think that, convener, we should look at commissioning our own independent research on this. This is a very important point about the future long-term security of tax income in Scotland. We have powers as a committee to commission our own independent research, and we should ask the class to prepare a draft paper on that, not to duplicate the work that is being done but to build upon it. It is fairly fundamental to our role in looking at the revenue side. We tend to look more at the expenditure side, but looking at the revenue side answers to those questions are pretty fundamental to deciding what the right policy is. It is certainly an argument that Mr Neil has been deployed by a couple of parties in Parliament recently, so it would be interesting to look at. Do the witnesses want to add anything to it? Only to say that it absolutely is a central part of the powers that are currently being devolved to Scotland from April this year. I have published a report today that looks at the roles and responsibilities and preparations that are in place. It is clearly a matter of discussion in the finance committee and other places, and we would be happy to help the committee with whatever work it decides to take forward in this area. Thank you very much. We will consider that and take it forward. Do members have any further questions for the witnesses? Or Mr Neil, did you have anything further? I just want to guarantee that Liam will not be flitting south of the border to dodge the tax. Monica Lennon It was just something that Caroline Gardner said in her remarks about the people who have afooked in both camps, including Scotland and other parts of the UK. How much of an administrative challenge is that going to be where people become more aware of the differences in some of the bloop holes, if you like? How much of an administrative issue will that be for HMRC, for example? We have pointed to an R report from 217 onwards about the compliance activity. That HMRC will be developing further in respect of the question that you have just asked. It is that assessment that I talked about in terms of how you define who is a Scottish resident for paying tax. This is an on-going strategy that HMRC will be developing in earnest from now on, particularly as the threshold has advised. They are well aware that that is an issue that they need to keep on going and keep on top of, because it does the risk that we have just been talking about earlier on. This is something that has been touched on by a few members in the committee. From your report, it is key that there is this issue about letters, communication and identifying potential taxpayers in Scotland. Is there going to be a loss to the public purse in terms of the communications perhaps not as good as we would like it to be? It is not just about the letters that have been issued. There are already in place that HMRC will certainly have other ways of telling or informing taxpayers as to what they need to do. It is not just about that letter. There will be other mechanisms there as well. That is the whole point. If that database is not kept up-to-date, if people are not aware of their responsibilities for informing the tax authority, the consequences are that, as we have just talked about in terms of those block grant calculations, they will be misaligned to the taxpayer community here as well. That is the risk that we have talked about. If you do not have your database right, you will not have your Scottish rate of income tax calculations correct, and that does have how much and by what degree that will lead to national economic calculations being adrift, exactly what we have talked about here today. As one of my colleagues said, a lot of that responsibility falls to the taxpayer themselves to keep their addresses up-to-date. Is that the same as operates currently across the UK on income tax or is that new? It is the duty of all of us, whichever side of the border we are on, to inform the tax authority of our change in circumstances. I mean, should we be—sorry, Carline Gardner. I was going to just add to what Steve Corbishly has said, that that is absolutely right. It is a duty on every UK taxpayer, and it has not really mattered in the same way until tax powers, income tax powers, were devolved to the Scottish Parliament, because we all paid the same amount of tax wherever in the UK we lived. That is now changing. That is the reason why the identification and maintenance of the taxpayer base and then HMRC's compliance activity are central to making sure that people pay the right amount of tax and that the Scottish Government receives the right level of revenue. Is it possible? A couple of my colleagues touched on this, that if somebody had a house in England and in Scotland that they could register to be taxpayers in England, I understand that it is residency for most of the year that was contained in your report, but how do you check that? It goes back to some of the answer that I gave earlier on. You test it by what—the reason is why you determine what your home is. If your family is living in England and you work in Scotland, then you could prove to HMRC that you are an English resident and therefore an English taxpayer. If the reverse of the case, then clearly a taxpayer would be proved to be a Scottish resident. At the end of the day, if HMRC needs further evidence, they will go on numbers of days. The numbers of days that you physically live in one country or another will determine whether you are a Scottish taxpayer or a rest of the UK taxpayer. If the CNAG's report talks about HMRC's compliance strategy for Scottish income tax, that is a key part of that. Given current tax policy with the higher rate threshold being frozen in Scotland while it rises in England, I suspect that their compliance activity will be focusing on people who fall into that band and particularly higher net worth individuals. Committee members may have heard of a group of people referred to as Willys who work in London and live in Edinburgh. Those are the people who have perhaps got the greatest opportunity for manipulating their tax status as tax policy diverges. I think that those are the sorts of groups of people who HMRC would be expected to be focusing on to make sure that the residency test is being applied rigorously and consistently. On that as well, we have pointed to 230. Those categories of people are very high. Regardless of where they live, they already have an allocated relationship manager who deals with their personal tax affairs to make sure that they are reporting the right tax in the first place. Of course, residency will be part of that assessment as well. Are you both satisfied with the HMRC's communication policy that we are going to be able to properly identify taxpayers in Scotland? We will test that and we will see the outcome of all that in the work that we do in 2016-17. We have seen the line of travel. We have reported on that line of travel. We can see how HMRC has continued to refine that database. I am giving a cautious yes, but we have to test that in terms of what we now see as the outcome of the first year, covering 2016-17. A cautious yes? Caroline Gardner. I have already confirmed to the committee that I agree with that it can provide assurance to the committee about the CNAG's findings. In my own report to the committee in paragraph 20, I make the point that the Scottish Government now needs assurance that HMRC's taxpayer identification strategy is fit for the future, having been through this initial phase, given its importance in making the overall devolution of income tax work for the longer term. Indeed, that is the page that I am looking at. I take from that answer, Auditor General, that you think that they are going in the right direction, but might need to be tightened up in the future to assure the Scottish Government. You have heard from the CNAG's report about the difficulties that were encountered in 2015-16. We are assured by HMRC that they have now been addressed. We will be looking for that assurance, and I will be looking to make sure that the Scottish Government has the controls in place to make sure that it knows that its taxpayers are identified and that tax is being collected as planned. Colin Beattie. There is just one question that came to mind. I am sure that it came up in times past. People working in the oil industry, where are they deemed to be resident for tax purposes? It depends on where their home is. It is that residency test. Again, they will have a home, and that will be where they will be assessed as a resident. In making this assessment, it is clear that people have got to pass the first hurdle of whether they are a UK taxpayer before you start to think about whether they are an English taxpayer or a Scottish taxpayer. That might be relevant to that question as well. Any further questions from members? I thank you both very much indeed for your evidence this morning. I am now going to move the committee into private session and suspend until 10.35.