 Good morning, and welcome to the 14th meeting in 2017 of the Finance and Constitution Committee. I have to honour our agenda to take evidence on the Scotland Act implementation reports, and I welcome to the meeting Derek Mackay, the Cabinet Secretary for Finance and Constitution. Mr Mackay is joined today by Aidan Griswyd, who is the Deputy Director of the Fiscal Responsibility Division in the Scottish Government. I welcome both of our witnesses to the meeting and remind our colleagues to keep our phones in the appropriate mode at the same time, and I ask the Cabinet Secretary to make a short opening statement. Convener, I think that because it's a technical report, essentially, on the implementation of powers, I'm happy to just immediately go to questions between myself and official. Okay. As you say, most of it's technical. On obviously the fiscal framework that was agreed by the two Governments, quite deliberately less some issues unresolved and to be resolved over the course of time, such as measurement of Scottish VAT, the classification of Scottish income tax payers, the establishment of a system to enable the devolution appropriately of the social security system to take place. It would be useful to know from you, Cabinet Secretary, for an opener, what you see is the biggest challenges yet left to resolve. Are you happy with the progress that's been made, and can you see any difficulties or challenges that you think you should meet the Government aware of? I'm sure that you've done that already, the committee aware of. Thanks, convener. Yes, I've been setting aside budget issues that we've had debates over and determination on that, and of course there'll be on-going debates around budget priorities and policies commitments, and other debates around powers that we might not have on the actual implementation of the acts. I've found that the levels of engagement are positive and constructive, and there is genuinely a willingness to resolve the technical issues from agencies, departments, around what's been agreed through the Smith commission, the fiscal framework and the relevant legislation. I suppose that some of the challenges around implementation, if you take social security, is such a huge task in heriting the existing systems in the UK Government in relation to social security. Some other elements might be more straightforward, such as air departure tax coming from APD, where Revenue Scotland is engaging. On the big challenges, it is the scale of some of the transfer mainly around social security. If you take, as you've referenced, income tax, there was the issue about classification of all-Scotch rate taxpayers, which I'm confident has been resolved. There are processes in place to ensure that we're identifying all-Scotch tax without the matter for HMRC, of course, but we want assurance that that's being delivered. At the practical implementation level, that's the big challenge. The safe and secure transfer of social security powers is to ensure that we have a system that's fit for purpose and is ensuring that people are supported. That's all been carefully planned out by the community's secretary, of course, and at a political level when there are issues, whether it's the transfer of resources or detailed methodology. There's official constant dialogue in working groups, leading to the Joint Exchequer Committee with Treasury ministers or engagement with the Secretary of State as appropriate or for social security specifically. There is a specific committee for that as well with our engagement. My general approach is to be constructive and positive in the delivery of the implementation of powers, but the risks are around transfer, the scale of transfer and the understanding that there are bureaucracies and systems involved as we are designing our replacement systems. Of course, that's the end aside from the other challenges that we'll face around budgetary pressures. As the new powers come to the Scottish Government, do we make sure that we have people in the right place to be able to deliver those new powers? Cabinet Secretary, I know that some of my colleagues will pick up on issues around social security, VAT and income tax issues and taxpayers. To do with one of the smaller taxes in terms of your basket of powers, there's been a bit of speculation and depressed in the last couple of few days about LBTT. I can see that I've not read beyond the headlines on that, so I'm not up to date on what it's really all about. It would be useful to hear from you what your perspective is on the LBTT situation in Scotland. I think that the position would be, as it was last time that I appeared at the committee and we discussed it, which is that we obviously make forecasts and it's not an exact science in setting out our forecasts. There have been economic issues around the Northeast, so overall in Scotland the number of transactions and the value of those transactions has maybe not increased as earlier forecasts would suggest. I hope that we've had debates here and elsewhere around the levels of taxation. There's no evidence that the levels of taxation are impacting on the market. When I've looked at the composition and the structure of tax take, it's operating similar to as before LBTT, although we feel that we've encouraged the end of the market in terms of first-time buyers and so on. The figures on the public domain around forecast and out turn, if you take both years together, our forecast compared to out turn for financial years 2015, 2016 and 2016-17 was within £13 million of a difference, having raised £13 million less, but that's a 1 per cent difference from our forecast to actual out turn. I'm saying within that that, of course, forecasts are adjusted based on the best information that you have at the time and there's no doubt that there's been an impact in the economy, particularly in the Northeast because of the oil and gas downturn and that has affected overall forecasts. We wouldn't be the only Government that says that this tax can be quite volatile and difficult to forecast. UK Government would say the same or OBR would certainly say the same. Indeed, they've said in relation to member UK equivalent tax that's stamped duty land tax as one of the more volatile sources of receipts. In line with that, we've revised SDLT forecasts proportionately more than any other major tax. Our forecast for the two years that's been in operation has been close, as you say, within 1 per cent of a difference. It is a volatile tax, difficult to forecast. It's easier to count what's already been received as opposed to looking ahead. Therefore, we reflect on all the economic determinants to assess our forecast position. Of course, the Scottish Fiscal Commission taken up the statutory function will determine the models that they want to use to forecast going forward and, of course, we'll be reliant on them going forward. I hope that that assists, convener. That helps to understand the Scottish position, but, obviously, this is a relative exercise, because it depends on what's happening north and south of the border. What's the up-to-date situation in the rest of the UK? In the rest of the UK, over that same period—I've covered 2015, 16, 2016 and 17—both years taken together, as I say, the forecast was £919 million. The actual outturn was £906 million, so that's a difference of £13 million—1 per cent of a difference between forecasts and outturns. The equivalent tax for the remainder of the UK for SDLT, their difference is £2.9 billion, so that's a 22 per cent difference that they were out by. That makes the point that OBR was saying that it is a more volatile source of receipt, and that forecasts are just that forecast. When you contrast and compare what we were able to set out by way of our forecast, an actual outturn was 1 per cent of a difference. The UK Government was 22 per cent of a difference, £2.9 billion, and I think that it shows that our forecasts were solid now. I don't want to speculate on that. That means that we'll always be within that range in the future. You don't know, but, of course, we're very reliant on those forecasts and setting out our budget position, but that's the difference between our forecasts and what the OBR had forecast. That sets the scene, cabinet secretary, and I know that other members are interested in this area. Murdo. I want to come in just to get some clarity, if I can. The note that we got from our adviser suggests that, for the year 2016-17, we're looking at LBTT liability in Scotland of £494 million. The block grant adjustment figure for that year, for both LBTT and landfill tax, is £600 million. If you add 494 and 114, you already exceed that. The question is, do we know what's happening elsewhere in the United Kingdom for that year, 2016-17? Do you think that the block grant adjustment is likely to be higher or lower than £600 million quartered? It comes from, you know. I think that I can comment on that. At the last budget, the £600 million was a sort of agreed figure. Since then, when we get to our outturns, the December budget had £641 for LBTT and landfill tax together, so that was an update at that point. In March, we've had, obviously, the OBR and the actual end result stats from HMRC on SDLT, but we haven't got the block grant adjustment figure relating to that yet, so we'll get that in a few months' time and that's something that the UK Government produces. Then we look at that and then we set it alongside what would be, at that point, a more final definitive figure on our own receipts because of accruals and the like as well. That's the point where you can make the calculation. If you like, it depends how you look at it—a deterioration of the block grant position at the December budget because it was £646 at that point. Since then, we'll just need to work through the numbers and wait until what the UK Government comes forward with in terms of that final estimate at that point. I understand that correctly, but the block grant adjustment will depend upon the relative strength of the collection of stamp-duty land tax elsewhere in the UK. Do you have a sense, as of now, of how that is looking or is it too early to see? The only sense that we have is that the final figures were slightly higher than the March budget forecasts by a few hundred million, so, to be precise, we've got £11 billion, £413 million was the March budget forecast, and a final out-term was £11 billion, £713 million. However, because of the complications of the block grant adjustment methodology, we don't want to ask the UK Government to crank the handle on that and confirm those figures in due course, but that, hopefully, gives you a little bit of an indication of the type of variation that we've seen in the past few months. £11.4 was what the UK Government had said in the spring budget. Is that right? Yes, we are. That's right. The red book's got £11.6 million, but it's okay, and the out-term is £11.7 million. For that order of magnitude, what kind of difference would that make if those numbers stood up? What kind of difference would that make to the BGA? Well, you'd have to do the Sarmys. It's £300 million, roughly, isn't it, over £11 billion or so? You can get a sense of scale from that, but it's, you know, well under... Oh, it's about 2.5 per cent or something, right? So, that would be a proportional impact on that £600 million, roughly. Yeah, and I think this is where we can't be too definitive, because things like those final... Sure, sure. But exactly in all those sorts of things. Since you're a scale, it's not a huge proportionate impact. But why is that if it's somewhere between 10 and 20 million, kind of number, whatever, right, go and take? And that would reduce the BGA then? By that amount? That would increase it, if all of that flows through. But as I say, I think it's too early to say, you know, the scale. Okay, okay, but as you say, the turn number you've got for the BGA is what you said, 640-ish? At the budget, the figure was 641 for SDLT and... Together. ...and landfill tax together. Okay, okay, okay, that's fine. I would, yeah. So, go over to Adam, social security issues. Thank you, convener. Good morning, Cabinet Secretary. I wanted to ask you some questions around the on-going project of social security devolution. I wanted to start with a figure that the Minister for Social Security gave to the Parliament, I think, last week in her statement on the social security agency, where she said that the estimated annual running costs of the social security agency are £150 million. She wasn't able to say anything about where that figure came from, or how it's composed, or how it compares with the annual costs of delivering social security in Scotland now. So I wondered if she could help us to understand that figure a little bit more, its provenance and its various component parts. Well, that would be for that portfolio to explain the individual costs of the operation of the social security system. What I will negotiate with UK Government is the transfer of resource. It's recognised that it won't cover every cost, but it's part of the fiscal framework agreement on that transfer of resource, which has happened incrementally over the period of this parliamentary term. So it would be for colleagues in communities to go into further detail about the arrangements for the agency and expenditure to see my engagement on this element as the implementation of the new powers to ensure that we have the transfer of resource. Necessary functions from the UK Government. Do you understand anything about where that £150 million figure came from? The figures will be understood by the minister and cabinet secretary. As I said, I won't have the detail right now on that operation of the agency, but my prime function here is to ensure that there's adequate transfer of resource and function from the UK Government. In terms of the question of transfer and resource and cabinet secretary, a couple of questions around that in terms of just trying to understand where exactly we are in this journey and our preparedness to take on responsibility for the delivery of devolved welfare and social security benefits in a couple of years' time. First question is, have Scottish claimants of the devolved benefits been identified by DWP? We know that there have been some issues around identification of Scottish taxpayers, but we also now need to identify Scottish recipients of devolved benefits. Has that been done? Again, I would need to refer you to the cabinet secretary on that level of detail around that. That's not something I would specifically deal with. You don't know whether any data on this has been shared with the Scottish Government. I'm sure that the cabinet secretary and Gene Freeman, the minister, will be able to go far more detail about the arrangements for benefits and payments than I would be able to now. I thought that you said in your answer to my first question about the £150 million that it was your responsibility to ensure the smooth transfer of resources from the UK to the Scottish Government to prepare for social security devolution. I think that those are questions about that process of transfer. The UK Government is not transferring the exact cost of administering the system, what we have is a deal around the transfer of resource. That's what's been negotiated through the Joint Exchequer Committee, and that's what I'm undertaking, that political deal, as to the transfer of resources. The lead cabinet secretary and minister would deal with the arrangements for social security in Scotland, and the delivery of those payments. What I have engaged with is the implementation of the powers. The resources transferring to us doesn't necessarily absolutely match the cost of delivering the new powers and services. That was a political deal. Is it your understanding that the cost of delivering the devolved powers will be greater than the cost of delivering those welfare benefits is now? Yeah, that's very possible. Can you help us to understand the nature of the additional cost? We won't know the full extent of that yet, knowing that we'll be transferring something of 15 per cent of social security spending, of course, 85 per cent remaining at Westminster, so we're taking part of system, we're taking part of payments into Scotland, creating those new systems and that bureaucracy to be able to deliver it. There will be an on-going monitoring analysis of what that will cost, and we're setting out how we want to design the system, some of the principles around it. The devalue of course of transfers is £2.8 billion, but it's too early to say at this point the exact cost of the new social security agency is delivered over the term. Okay, but delivering that £2.8 billion worth of social security benefits will cost more after devolution than it costs at the moment before devolution, is that correct? I'm sure that it's quite possible, because we're taking part of a system from the UK Government that it'll cost more than what we're spending at the moment, because we're administering very few types of these payments. We anticipate the cost being more than what has been agreed in the fiscal framework just for social security. Just help me to understand this. If the mere fact of devolving the responsibility to deliver these benefits has an additional cost, how does the no detriment principle apply to that? Is that a cost that will have to be borne by the Scottish Government? Well, essentially, of course, we want to be as efficient as possible. No detriment applies because of essentially the wider deal on the fiscal framework that there's no detriment overall to expenditure in Scotland as to how we use these powers. As I say, we could have continued a negotiation with the UK Government—I wasn't partied to it—but we believe that we've got an adequate transfer of resources to be able to deliver those functions, and that's what's been agreed with the UK Government. However, the additional cost of delivering those social security benefits in a devolved setting over and above the cost of the way in which they're delivered now is a cost that will have to be borne by the Scottish Government. Yes, well, essentially it's a devolved function, yes, so we're paying for it. A final question for me on this, if I may, Cabinet Secretary, are we, from where you sit, given your portfolio responsibilities, on course for the transfer of operational responsibilities by April 2020? All of them, from my point of view, I believe that the necessary work streams are under way, and there's a commitment to deliver on both parts of Government, yes. So I understand that there's a commitment to deliver, but are we on course to deliver? In my view, I've seen no information that suggests we're not. I took that as a yes. The Cabinet Secretary nodded. I'm trying to express my confidence in the engagements that we have that we're going to deliver the agreement over the course of this term of Parliament in the way that we'll set out, yes. Thank you. James. Thanks, convener. Just following on from that, I'm interested in the timelines around us. Obviously, it's been transferred in terms of two tranches, so if you take tranche one, just as an example, in terms of discretionary housing payments, can you maybe then just talk us through the process for transfer and the key timelines in relation to discretionary housing payments? Again, the Cabinet Secretary would be able to go into more detail on individual payments, powers and elements of that, rather than myself. I've engaged with the top-level agreement around transfer of power and resource. If you want further detail, I would encourage you to invite Angela Constance and Gene Freeman to go over the specifics. I'm sorry, I wasn't trying to be awkward in any way. I thought I was asking a reasonably straightforward question in the sense that the note that you've provided says that tranche one was commenced on 14 July 2016, and it gives a list of the sections that cover tranche one. I'm just looking for an indication—not an indication—a more specific answer in terms of what is the process for that transfer and what are the kind of timescales when and the way. I'll ask Aiden Grewswood to come in with detail. Yes, so specifically, Mr Kelly. Discretionary housing payments, so that takes effect and takes responsibility for that from April 2017. Some of those bespoke discretionary payments are taking effect this year, and then there's a range of benefits which DWP currently administer, which are more complicated in terms of delivery. Again, it would be for security colleagues to give more detail around this, but essentially that requires—it's a much more complicated process of transition and work with DWP to ensure that those are delivered successfully by 2021. Just going back to tranche one, which lists—I'm just looking at it—it looks as if there's around 10 there. You said that discretionary housing payments have effectively come into effect April 2017. Are you saying that the others have got different timelines on them, or are they April 2017 as well? Well, it's the powers overdoing this, and it's discretionary payments through local authority, so the housing payments will take effect from then, and then there's extra powers on top of that over the same timescale. To give another example, what about, for example, powers to create other new benefits? Has the legislative transfer been affected for that? It would be new legislation to provide new benefits, such as the social security bill that's coming forward that will lay that out in more detail. I suppose that the distinction that's being made is between, as you said, discretionary or new benefits that don't suffer the consequences of complexity of DWP administering at the moment, and that's a smooth transition process, so that's why those were selected to go earlier. Okay, in terms of these discretionary payments, the transfer has been made, are the systems in place to allow discretionary housing payments, as an example? Are the systems in place to allow that to go ahead? Essentially, that works through local authorities. My understanding is that it works through local authorities anyway, since it's using existing systems through discretionary housing payments. So, just to be clear, you're saying that, at this minute and time, local authorities now have the power to make discretionary housing payments? It's about who's responsible for this, so it's about the authority being passed to the Scottish Parliament rather than the UK Parliament. Yeah, but I'm not trying to be awkward. You said that local authorities had the power to make these discretionary housing payments. All I'm trying to establish is if they got the power now to do that, there is still some work to be carried out to ensure that that happens. No, so they currently have the power in our making discretionary housing payments. It's more about the devolution of responsibility going from UK Government to Scottish Government and therefore the funding around that being determined from Scottish Parliament rather than UK Parliament. So, is the funding in place for it, then? Yes. Right, just going on then, sorry, convener, just going on to change 2, which will be completed by April 2020 in terms of the executive, both legislative and executive competence, but there's a line in there that says thereafter agency agreements would be put in place as necessary. Does that mean that there's further work required after the executive competence has been transferred in order to take effect to the sections 22 and 23? Yes, that's my understanding, that's correct. So, it provides exactly the enabling powers and then judgments will be made working closely with UK Government around those specific tranche 2 benefits as to when those would transfer, making sure that there's a smooth transition and the interests of benefit recipients are fully protected. But you've got the enabling legislation in place to give that legislative competency to the Scottish Parliament to make, to move quickly when it deems that it's reached that point. So, it's fair to say that, obviously, there's been concern expressed that this is going to take four years, but it's actually going to be actually greater than four years by the time all the appropriate systems and processes are actually in place in the agency agreements. I think you're getting into levels of detail, which is entirely appropriate and fair, but really for the community's portfolio, if you want to probe it further, I think that's fair, that the detail around implementation would be for the cabinet secretary and minister. So, I'm happy to arrange that, convener, if you want more detail on that. I know a statement that's been given to Parliament. We're further outlining the role of the social security agency, and we do have a range of commitments as to how we would use the new powers, but if you want more detail on that portfolio level, then that can be given. I'm not going to ask any further questions, convener. I would just say that, obviously, there are concerns as to why this takes four years. I understand that it's complicated, but four years seems an awful long time, and from what's been said, it's actually going to take greater than four years. I think that that is a matter of concern. I don't think we're saying that. Obviously, we want the safe and secure transfer of powers from the UK Government to Scotland to ensure that people get payments that are entitled to, and we have that flexibility to do things differently. There are arrangements in place in being formed as to how the social security agency will be delivered. Of course, there's engagement with local government as well. My point is that, if the committee is, of course, entitled to do once more of the detail around what we're proposing around the agency and payments, then it's more appropriate to get that from Angela Constance and Jane Freeman. I wonder if I could just go back for a moment to the identification of Scottish taxpayers and the numbers involved in that. We know that HMRC initially failed to identify about 420,000 Scottish taxpayers, and that has been resolved fairly quickly. You said in your opening remarks that you're confident that that's been resolved, but what form of assurance is it simply reporting sharing between one Government and the other, or do you have access to the data, as are data sharing going on, and who's verifying the data on an on-going basis that we've got accurate numbers today? It's fair to say that we're concerned by that classification and identification issue. It is a matter for HMRC to provide that service to us, but they've gone through it quite methodically to be their confident that they've identified all Scottish rate taxpayers. My understanding is that there has been some independent work as well to look at that third party analysis to make sure that we're confident that that's been arranged. There's some other processes that you can undertake, and we'd encourage taxpayers to check their code as well, but we are confident from the engagement that we've had through officials that they've identified all Scottish rate taxpayers. We're also going to have a service level agreement to make sure that we've got an adequate level of service as well, but we don't believe that there's a large scale or any outstanding Scottish rate taxpayers that are now not identified. Just to build on that, there's always movements of the tax base. There's a particular issue around the 420,000 and HMRC acted, they did a short-term fix and then they've run an IT system to catch those 420,000 taxpayers. It's all about reasonable efforts. How much can you reasonably capture at any one point? I guess that's the point. At this point, we're very assured that they're making reasonable efforts to capture that taxpayer dataset. It's always one that changes over time because you never know. There is no ideal set for you to compare it against, hence the point about making comparisons with third-party datasets and the like. It's not necessarily because the third party datasets are fuller than what HMRC has, but it can flag anomalies or issues that they can then address. There was the big issue of the 420,000. There's issues around things like missing addresses and the like, much smaller scale that they're working through. Issues about postcodes here and there not being captured. Each of these is iteration after iteration. It's a continuous improvement process. We stay very closely to that to get assurance ourselves that that works being undertaken. We have regular review meetings with HMRC to understand what the current position is, what work they're undertaking to improve the dataset. We did this on our own offering, looked at, for example, the Scottish Government payroll to see whether that flags out issues that they can then address. They've also done the same with other large employers. They've undertaken communications activity last year, and they're looking again at how that can fit into communications campaigns around the individual personal account of HMRC, which is something they're trying to promote just in terms of efficiency and being a positive thing for taxpayers themselves to see their own tax position, but rolling that out will improve because, by setting up an account, you update your address. The emphasis is always on, there's only so much you can go in terms of the datasets. Ultimately, people have to update their address, so the challenge there around just ensuring that as many people do that as possible. What they're also doing is developing a compliance strategy on the back of the rates that were set in the budget, which should be completed fairly shortly in terms of activity that will be undertaken to ensure that, again, people aren't shifting that there isn't activity that would lead to tax avoidance. There's a range of measures. We have regular review meetings with them. We've got assurance from that that they are taking this very seriously. There's a continuous improvement process. Separately, we've got, of course, National Audit Office and, indeed, Audit Scotland recently reviewing that too, and they've given evidence to not this committee but the recently public audit committee on progress there. Okay. We've got two papers in front. There's one from yourself, the Cabinet Secretary, and one from the now former Secretary of State of Scotland. In your paper, Derek, in paragraph 41, it says, sorry, how we can come here now. I'll just continue in post, even if there's an election, just for clarity. That's very helpful, because I think we should be asking them a couple of questions, if you let me ask my question. Derek, in your paper, HMRC tells you that they have 2.6 million Scottish tax payers identified as paragraph 41 in your report. When you look at the current Secretary of State's report on paragraph 10, he identifies that their 2.45 million letters were issued and then 420,000 were missed. So that total is 2.85 million. That's a difference between his assessment of the number of tax payers being 2.8 million and HMRC telling you in your paper that it's 2.6 million. That seems to me that there's still quite a big difference there in the number of identified tax payers between the two reports. We can come back to that in more detail if need be. There's also just a difference between the 2.6 million, which is HMRC's understanding of how many tax payers there are out there, and that's what flagged the 420 originally. There's also the further question of the extent to which people who were flagged as being tax payers ultimately paid tax because of through self-assessments, which led to a further change in the number. I suspect that it might be those figures being interpreted differently and laid differently that have come to that mismatch, but we can certainly clarify that, if that would be helpful working with IUK Government counterparts. Thanks for asking that question. I do think that it's good enough to have a situation where there's different interpretations being put on the number of tax payers we have. We need some assurance in that in Scotland on exactly how many tax payers we're going to have. I'm glad you raised that. I hadn't spotted that in the reports, really, so we need to... Maybe something we'll have to consider following up on, because if we haven't got a solid base to start with and there's disagreement about what that base even is, then we've got an issue in Scotland about not how many tax payers we have. That's just no good enough. That was going to be my next question. What is the total of Scottish tax payers, as we understand it, and as the Secretary of State might understand it, and perhaps the committee might consider inviting the Secretary of State in to answer the same questions that we're asking our Cabinet Secretary, what is the actual correct total? If we're dealing initially with an anomaly of £420,000, now we have two papers in front of us that still have £270,000 of a difference, I think we need to be scrutinising this a little bit more carefully, I think. Fair convener, we both, both Governments have the same source and that's HMRC, who we pay to carry out a function they deliver a service to us. They identify the tax payers' follow-through on compliance and, of course, collection, so we will happily engage. I can't answer for UK Government. There is no policy dispute here, but we can certainly explore the point, but HMRC is the same source for both. It may well just be a descriptor at a point in time, but, as I said to the earlier question, I'm satisfied that HMRC have undertaken the efforts to ensure that they've identified Scottish tax payers and that's who we're collecting tax from. That earlier issue that had been identified in 2015 has been addressed, but we'll certainly take up the point and the narrative about the difference in position. As to the Secretary of State, but being here to answer for his report, that's, of course, a matter for him. First of all, a matter arising from Mr Coffey. Mr Greaves, would you talk about the income tax relying on people notifying HMRC of a change of address? Of course, the National Audit Office has said that there remains a risk that people won't update their addresses and then you get the anomalies identified. How is that being addressed, or how are you addressing that, given that, as I understand it, there's no further spending commitment on communications to people? Also, how do we ensure that there is no artificial manipulation, just as you mentioned, if perhaps the tax rates diverge? Well, it remains a risk, as you say, so it's an ongoing issue across the piece trying to get people to update their addresses and encourage that to happen. So, as you say, there was the communication campaign previously and I think the focus is now shifting towards a more generic campaign around individual accounts with HMRC and I think that presents an opportunity for HMRC considering in terms of the extent to which that could be intensified in Scotland in order to have a double benefit as a consequence of that campaign, both in terms of encouraging people to use online accounts but also through that process updating their address. So, that's part of it, that's a kind of communications strategy. There's also a lot of work going through their large business unit, which is actually working directly with employers, which is a means of raising awareness of that, so there's a link through to that as well. Then, there's a compliance strategy coming back into your last point, which I'm not liberty to say, because it hasn't been completed yet, but HMRC are developing a compliance strategy for Scotland on the back of... Just to make one point, of course, just for clarity and understanding that the tax position is based on residency, of course, and that's a major driver in where you're paying a tax, so just for awareness of committee, that's an important point. When they focus on compliance, that's what they'll look at, issues of residency. So, your point around just flipping addresses, clearly HMRC, we would focus on that and the evidence of where someone actually lives most of the time. I don't want to go into the detail, I don't want to describe ways in which people can avoid paying tax, of course you wouldn't expect me to, but residency is the key issue here. I should say that there are all compliance strategies that the sums at stake here in terms of tax differentials are far smaller than some of the sorts of the scale of issues that they normally have to face in compliance in terms of very high-rate taxpayers and extreme-wealth individuals and opportunities, incentives there to potentially avoid tax. In the scale of those numbers, a huge incentive is much bigger, so they tend to be more focused inevitably in terms of targeting resources on that level, high level of compliance rather than necessarily sort of small variations in tax rates. Just staying on income tax then, the HMRC running cost of the Scottish rate of income tax is £1.5 million for this year, and they estimate, I think I'm right in saying, that that running cost will increase to about £5 million if the Scottish rate of income tax is different to the rest of the UK. Can you just explain, first of all, why that is? That's a big jump. That's a maximum figure, and actually, as it happens, it depends on the scale of the change that they're talking about, so there's a contingency figure. Next question might be useful then. What I was going to go on and say is that is there some kind of de minimis on this? How different does the Scottish rate have to be to having, for you to incur an extra £3.5 million in cost, or is it incremental? It depends on the scale of the change, you say. If there was a sort of wholesale change, different bans, different rates, that would be a whole different magnitude of potential cost implementation. If that was late in the day as well, so that was sort of just before the financial year started, that incurred a greater cost as well because of the intensive work that would need to be done and they'd have to do extra work just to work through the systems and update them, which they wouldn't necessarily do through businesses' usual practices, as against a small change. A small change will, in advance, be in very small costs. Is it essentially cost based on what the policy issue is, what the difference is, and then would return to us with what the cost is? Right, but if there is, just for clarity, if there is a small divergence in the rate of tax, then there will be a cost incurred, which presumably needs to be budgeted for. Yes, we pay HMRC to carry out a function that costs that and return to us, yes. UK government changes taxes at a cost. I don't think there's any great surprise in that stuff. That's difficult. VATs even more challenging, I think, from what I've been beginning to understand and Marie was going to pick up on that. I wanted to ask if, just about the methodology involved in calculating the VAT assignment, now that it's going to be such an important part of our budget going forward, I understand that, in the past, with the GERS figures, an estimate was sufficiently robust, but we need something a little bit more accurate and robust going forward when it's going to be such an important part of our spend. As members will know, the agreement is not that it's actual spend or receipts in Scotland, it will be just that assessment, an estimate that leads to the assignment of the proportion of VATs. However, we are still working on the methodology as to how that will be applied for 2019-20. Edin, can you maybe say where that is? Yes, that's quite a bit of intensive work going on between our analysts and HMRC analysts around developing that methodology. It draws on existing work in terms of a regional model of VAT and enhancing that. Part of that is basing it on survey data, using expenditure surveys to get an assessment of what expenditure is in Scotland by different categories of goods and therefore able to drive an overall estimate of assigned revenues through VAT. That's been enhanced by a boost to that survey, so it uses a UK-wide survey and is a boost that's been paid for for Scotland specifically in order to get a better sample side. That, therefore, helps to reduce the volatility of the number that comes, which is part and part of the consequence of using a survey, rather than the actual data, is that you need to reduce the volatility of that number. That will help in doing that. There's quite a lot of complex works involving particular sectors where there are issues such as reliefs and the like that you need to consider, or particularly, for example, the financial service sector, which is a very complicated VAT arrangement, so there's bespoke work that's going on to look at all of that as well. One of the concerns that we've had in the past is that using household surveys and extrapolating from that might not capture the increased spend that comes from tourism in Scotland, so tourism is a significantly greater contributor to the economy up here than in the rest of the UK. There's a lot of policy aims to increase tourism, so, for example, if we reduce their departure tax, we would expect to see an increase in tourist numbers and we would expect to see those people spending more money in our shops and an increase in VAT take. Are we going to capture that in the new methodology? Yes, I understand that there is a specific bespoke piece of work around expenditure by tourists that would be used as a bolt-on to the overall expenditure. As you say, household surveys wouldn't capture that, so it's exactly that sort of additional work that's being done to try and cover some of the potential holds and existing approaches that's being done. We've got more time on VAT given the timescales involved, but it's quite a big task to get that comprehensive survey. We've also got the transition period, too, built into the fiscal framework, so it doesn't necessarily just all go live day one. There's an opportunity to have a transition period where the VAT revenues are protected, but we're actually looking at the figures and understanding those and then able to come to a view about the appropriate deployment to that methodology going forward. Finally, I presume that, after 2019, if we Brexit along with the rest of the UK, there is an opportunity for VAT to be fully devolved to Scotland because it's a European law that was preventing the full devolution to Scotland? There is that possibility. I wasn't something that we've negotiated in the previous act, but that would be in the hands of the UK Government. On this element, on methodology, our recommendation hasn't been put to politicians yet because it's still been worked on by the officials, but the important point around capturing as best we can actual receipts in Scotland is what we want to achieve. I have a question in this area as well, but Adam has a specific supplementary question. On the very last point that Mary Todd just asked, as I recall it, Patrick Harvie was there too, he'll correct me if I'm wrong, but as I recall it, the Smith commission recommended the assignment of a share of VAT receipts in Scotland to the Scottish Government, not the devolution of a share of VAT to the Scottish Government because to devolve VAT would be unlawful in EU law, each member state being required to set only one rate of sales or consumption tax. As we are leaving the European Union, what discussions have there been between your officials and Treasury or between yourself and Treasury ministers about devolving VAT? We've engaged in a range of matters, but not specifically VAT other than having made the point around more widely on the impact that VAT has had on tourism or hospitality or refurbishment of properties on, so it's been part of the much wider discussion on VAT, including the emergency services as well. That's where my engagement has been largely around VAT. December called Scotland's Place in Europe, which included a quite lengthy chapter about additional devolution, but as I recall it didn't say anything very much about VAT. Is that because you don't want VAT to be devolved? No, I'd be more than happy if the Scottish Government was able to have control over VAT as well. I'm just telling you specifically what I've engaged with UK Government as you asked on VAT, which was those other matters. Given that you want it to be devolved, why haven't you had any conversations with UK Government about that in the 10 months since the EU referendum? As Adam Tomkins well knows, we engage with a maximalist position on the devolution of powers to Scotland, but I've also accurately answered your question on what specific conversations I had with UK ministers on VAT, which was around emergency services of course. Scotland's emergency services haven't to pay for VAT, not being able to reclaim that, which is a great loss to Scotland's emergency services, or other areas that would be supported in Scotland. The other parts of the continent would be supported if we could vary VAT rates. Of course, we would support the devolution of VAT to Scotland, but you're also correct that the deal that we're discussing here, the implementation of the Scotland Act powers, was around that, as you have accurately described, an assignment of a share of VAT, but based in estimates it's not the devolution of the power, but it would not surprise Mr Tomkins to know that, of course, we would like that economic leave on others to be transferred to Scotland under any circumstance. That's the difference, isn't it? If we leave the European Union. I was going to get there anyway, but since Adam has opened this up, I might as well slightly reverse the order of my questioning. I did, just before the meeting got under way, have a wee quick skim back over the Smith commission final report. I do it every time I need a laugh. The report itself doesn't go into a great deal of detail about the rationale, about the reasons why what was being recommended was recommended. Obviously, the European rules were part of the discussion. At the time, I made the argument that I was pretty unconvinced about the assignment of a proportion of attacks, because it doesn't give the Scottish Government or the Scottish Parliament any additional policy levers, any additional fiscal policy flexibility. Since that time, as the Government's thinking developed at all on whether there is any benefit to the Scottish Government in policy terms of the assignment of a proportion of VAT, what's the point of it? I suppose that there's a point around more accountability in that general economic growth and delivering more economic growth being good for the economy, being able to estimate what's actually received in Scotland contributes to public expenditure as well, but Mr Harvie is correct. We would rather have had, of course, the devolution of the power to be able to use it in a way that supports the economy and has full transparency, accountability and commander over that, but what is being implemented is what is agreed, which is the assignment rather than devolution or control of the VAT. It achieves greater accountability in terms of one part of Government policy, economic growth, but no relevance at all to other parts of Government policy. I'm making the point that that's one benefit of having that assignment based on estimates, albeit in Scotland, but I'm in agreement with Patrick Harvie. It would be better to have devolution and control of VAT because it would be used in a way that supports the Scottish economy and as part of a range of tools that the Scottish Government would have to direct at economy and fiscal decisions in a more appropriate way, rather than be able to determine the tax but simply receive what's estimated to be our share of it within Scotland. If nothing else changes, the assignment of this large chunk of VAT raised in Scotland and the block grant adjustment taking that money back again, if nothing else changes, it doesn't really affect the Scottish budget. It only affects the Scottish budget if there are changes over time, either in the amount of VAT that's raised in Scotland or if there are errors in the methodology that's used to calculate it. Can I ask you if the approach to the methodology is not really a fundamental change, it's trying to get more accurate, roughly speaking, the current approach. Is there any potential consequence of taking as the baseline year a year in which, for example, the UK's future relationship with Europe becomes clear in the publication of a Brexit deal and that there are economic consequences that affect the amount of VAT that's raised in that year? Is there a danger in taking that kind of potentially turbulent year as the baseline? I'll ask Aidan to cover that particular scenario in as far away as possible. The key thing is that it's working through the block grant adjustment, so it's always about the performance relative to the rest of the UK. As long as there was a shock, if you like, economic shock from any particular reason, as long as that affected both the rest of the UK and Scotland equally, there wouldn't necessarily be a particular impact on Scotland. If it was diversion, if it was a different type of shock, then that would be a consideration in terms of when the baseline would be set. These are sorts of issues that we need to be very clear on in agreeing, because on the one hand it could be quite straightforward in terms of methodology, but there are some fundamental issues around all of this in terms of agreeing the final set of arrangements with Treasury about how this will work forward and indeed advising Mr Mackay on that as well. Given the concern about one industry tourism, which has already been flagged up as particularly significant in Scotland as a source of that being raised here, which we haven't yet got a methodology for counting, there is the potential for that kind of asymmetric consequence. If there was volatility year to year, that's partly why they're getting it up and running, getting the transition years in place, giving us a good indication of the extent to which there is volatility there. Have you confirmed yet which years those will be? We haven't. When do we expect to know? We've got a JEC plan for the summer with ministers. Sorry? So a joint JEC, a joint committee for the summer, which will be informed by the joint working of officials. And you would expect that meeting to make the decision about the timetable? Not necessarily. There isn't very long to go. No, but I think that the point that you're raising is a valid one around the baseline and the methodology and then potentially impacting on that for all future years on where the baseline is set, particularly the point in the economic cycle. So I think it's a fair point to make to consider what we say as the baseline, the start position and the methodology. Mr Havie makes a very valid point about what that would mean going forward, depending on where we are in an economic cycle. So that's part of the consideration, as well as seeing through the spirit of the agreement. We are talking about a block grant adjustment that's almost half the size of the income tax one. Is that right? And we're already concerned about potential inaccuracies in terms of counting income taxpayers. If there was any kind of ambiguity or inaccuracy either in the methodology or a similar concern as a result of when that is implemented, we'd be talking about a very substantial role of the dice in terms of future Scottish budgets. Now I think that that's exactly why we have to be absolutely sure of the methodology, but also the point in the economic cycle that we set as the baseline going forward. And Mr Havie touched on the Brexit issue. We know that the independent analysis is that there will be an economic impact on GDP, on tax receipts, on inflation. Now we can't forecast exactly what that looks like, but it will have an important impact nonetheless. So I think it's a valid point to be clear on the benchmark point and the methodology to ensure that we don't disadvantage Scotland's budget. Thank you. I'll let you tongue a cheek. It does make big the question of detriment applies, no detriment applies to Brexit or not. That's a Sam question, but I wouldn't say. I think convener that there were no unknowns and no unknowns and then there was Brexit. You would deal with issues around the Scotland reserve. Yes, so in paragraph 72 of the Scottish Government report, I'll just read it out. You said that detailed arrangements for reporting, repaying, borrowing and the operation of the Scotland reserve are being or the process is being agreed at the moment with the UK Government. So I'm just wondering, do you know at the moment whether or when the arrangements will be agreed or are you able to sort of put that on the record now? I can cover what's happened in terms of actual practice around budget and we understand what our borrowing limits are, we understand what the Scotland reserve should look like. Of course, there is the issue where in the financial year we generate more tax and we had anticipated we put that in the cash reserve but we understand what our limits are. As to borrowing decisions specifically for capital, they'll be taken for the financial year that we're currently in much later in the year but they can go through some of the detail but that's where the budgetary position is and the functions that we're currently executed and the decisions that we'll take going forward. Again, from a budget point of view, we don't have any plans to use resource borrowing but Eden can see more. Yes, I suppose that the final range is more the technical details as Mr Mackay set out. The high-level principles are all there in terms of the caps. Daniel Capps, the aggregate caps on capital and resource, some of these agreements are more about the precise application of those. For example, things like the rolling four-year period of economic shock, how do we define that exactly, these sorts of things. So it's when you get into the detail just agreeing all of those so that we're all on the same page in terms of future arrangements but the high-level, there's no doubt about the high-level agreement stance. So if the detailed arrangements are not yet fully agreed, I'm just wondering what sort of scope there would be for this committee to scrutinise those sort of details when they are agreed? Well, we can return to the committee when each of these elements are agreed if that's helpful and particularly on the intergovernmental relations report to the committee what engagement I've had with the UK Government so we would normally inform the committee of that. So we can return on what has been agreed and then the committee can choose what it wants to probe further. Neil, do you have some issues around general economic issues on this? We've just touched on it there but in terms of obviously we're all aware that economic performance relates to revenues in future years but latest economic growth figures show that growth in the rest of the UK has been higher than in Scotland. Do you envisage that trend continuing and if so are you concerned for revenues in Scotland as a result of that? Well of course we want to support the economy, encourage economic growth and GDP growth and I don't necessarily want to rehearse all the politics but we know that Fraser of Allander Institute identified Brexit as an issue, there's an issue around consumer confidence. There are clear challenges in the north-east around oil and gas downturn and that includes the supply chain on shore. That said, we've been strong in terms of productivity on unemployment, female employment and youth unemployment so there are many positives as well and we've had a strong record and foreign direct investment and we're doing more around that in terms of our strategy on recalibrating the economic strategy and internationalisation and so on. So there's a range of actions underway to support the economy as well as the infrastructure spending and a whole host of other actions. Of course we'd be concerned that the last quarter's GDP stats will hear the next quarter's stats in early summer but from my point of view the budget has made a very positive investment to support the economy and we want to continue with that but as I said we've got a high level of employment and a low level of unemployment but the challenges to Scotland's economy I think have in large measure been down to oil and gas. Now there are some sounds of recovery and growth opportunities there and hopefully that will help with the figures going forward. Okay, thank you cabinet secretary and official for coming along today and giving us evidence on the reports. The next meeting of the committee will address stage 2 of the air departure tax Scotland bill and now close this meeting. Thank you.