 Good morning, and welcome to the 13th meeting in 2019 of the Finance and Constitution Committee. We have apologies from Murdo Fraser, who has moved amendments on the transport bill this morning. Neil Bibby may have to leave at some stage during the proceedings for the same purpose as we go through the morning. I remind members to put their mobile phones etc into a mode that they won't interfere and they better do the same in mind. The first item on our agenda this morning is to consider the Scottish Fiscal Commission's economic and fiscal forecasts published last week to accompany the Scottish Government's medium-term financial strategy. I welcome to the meeting Dame Susan Rice, who is the chair of the Fiscal Commission, Professor Alistair Smith, one of the commissioners and John Ireland, the chief executive of the Scottish Fiscal Commission. I warmly welcome our witnesses to the meeting this morning, and I wonder if Dame Susan Rice would like to make an opening statement. Convenor, thank you very much and thank you all for inviting us to give evidence again today. In addition to my colleagues who are here with me, I would just give apologies from our other two commissioners, professors Francis Breedon and David Ulf, who were not able to join us this morning. I'd like to give you just a bit of an updated, generic update. Since we last appeared in front of you in January, the commission completed its second year as a statutory body. I'm very pleased to say that we have now signed a memorandum of understanding with DWP on access to data and social security. I'm grateful to the civil servants at DWP and the Scottish Government who helped us with that. You'll remember in January that we had just then signed a formal MOU with the OBR, and we're now about to sign a revised MOU with HMRC. That's all really good progress from our perspective. As you know, our founding legislation requires that we're externally reviewed at the end of our second year. I'm very pleased that the OECD have agreed to carry out this review for us along with several international experts. The review team will be in Edinburgh in fact next week, and I'm grateful that a couple of members of this committee will be meeting them on Tuesday morning, so thank you for that. As you know, we publish our forecast twice a year to support the medium-term financial strategy of the Scottish Government that's in May and also their budget in December. The latest report contains our economy, tax and social security forecasts as usual, but what we've done a little differently this time is to put greater emphasis on an analysis of key issues facing the Scottish budget. Everyone's attention at budget time in the winter is focused on tax and spending plans for the year ahead. What's good about the summer, the most recent forecast, is that with the immediate pressure of setting a budget behind us, we can spend some time thinking about the longer term. Our forecast look ahead over the next five financial years. We also take the opportunity of adding an additional financial year at this stage to our forecast. This time around it was adding 2024-25. In our report, we highlighted two longer term risks to the Scottish budget. The first arises from the devolution of further social security benefits in April 2020, so that's in less than a year. Our estimate of the spend on social security next year is 3.5 billion pounds compared with the £447 million that we expect to be spent this year. The forthcoming benefits are demand-led. In other words, anyone who applies and is eligible must be paid. The Government will need to be able to manage, in year, any difference between the forecast and the actual spend. By way of context, the Government's entire spend this year on its justice portfolio, which covers police, fire, court and prison services, the entire spend in that area is £2.7 billion. You can see that £3.5 billion is a great deal of money. What makes this trick here is that forecasting the spend on new benefits to be administered in a distinctively Scottish way and possibly around different eligibility rules is by its nature much harder in the first few years when we don't have an established baseline to work from. The second risk that we highlight is the adjustments to the UK treasury will begin to make to the block grant, reflecting for the first time actual income tax collected. We estimate that these adjustments or reconciliations using the technical language of the fiscal framework will reduce the Scottish budget by £229 million next financial year and by £608 million the year after. The reconciliations arise from the use of two sets of forecasts at the time that the budget is set. Our revenue forecasts and the forecasts of the block grant adjustment are based on OBR forecasts of receipts by the UK Government. Forecasts, as we all know, are never entirely correct, but the budget has to be based on the best possible estimates of what will be raised and spent. I suspect that you'll be interested in why our estimates of the reconciliations are the size they are, and we can explore that with you. At this point, I think that it's worth saying just two things. First, as we said in December, considering the OBR's track record over time, we may see errors as large as 3.3 per cent or around £530 million in a Scottish context. That was just to give you a sense of what might evolve over time. Second, at the moment, our analysis of reconciliations is based on our and the OBR's most recent forecasts. During this summer, income tax outturned data for 2017-18 will be published. In our forecast evaluation report, we intend to present a detailed analysis of the actual reconciliations, which I think will be a fruitful discussion then. Returning to the Scottish budget, what's probably most important for the Government is their decisions regarding how any volatility in that number is managed. The Government can borrow, can use its reserves to help deal with these reconciliations. It also may have to consider adjusting its spending plans. Finally, I'd like to turn to the prospects for the Scottish economy. In December, I said that Brexit was at the front of our minds that has continued to be the case when we started working on our forecast back in March. We thought hard about how we would deal with Brexit. Our forecast is based, as it was in December, on a broad assumption of an orderly negotiated exit from the EU. We now, obviously, assume that exit will be in October rather than in March. The terms in which the UK may leave the EU are still highly uncertain, and we've made a number of broad-rush assumptions to capture a range of possible outcomes. A no-deal exit is not captured in our central assumptions, but is a significant downside risk to our forecasts. We followed the many twists and turns on Brexit as we put our forecasts together. We finalised our approach at the beginning of last month when the Government needed our final forecast to do their work. While things have moved on since then, most recently the Prime Minister's resignation, we still believe that our Brexit assumptions are a reasonable basis for our forecasts. Thank you very much for your attention. That was to give you an overview, and we now welcome any questions or thoughts that you have. Thank you, Dame Susan, for your opening statement. You rightly mentioned the forecast of income tax reconciliations. We need to get into that area first. Something is puzzling me about that. Despite Brexit, there is a slow projected population growth in Scotland, compared with the rest of the UK. Both of those elements are largely out with the control of oil and road. The Scottish economy is still growing. Unemployment in Scotland is lower than the rest of the UK. Our tax take continues to grow, and yet we end up through the forecast of the income tax reconciliations with less money. To me, that is not how a normal economy would work. There is something in that that we need to understand a bit more. That is pretty clear to me. For the benefit of the committee and those who might be watching the proceedings and who are interested, perhaps you could let us know from a Scottish Fiscal Commission's perspective why is that happening. A really important question. We have spent a lot of time considering how to articulate a response to that. Alistair, can I ask you maybe to kick off? Yes. I will have a go at that. The reconciliations process is a very complicated one with lots of moving parts, and it is very hard to keep your eye on getting all the moving parts lined up in the right way at any point, but I will do my best. The key starting point, as Susan said, is that when the budget is set—and let us take the 2018-19 budget as an example because that is where the biggest reconciliation is—when that budget was set in 2017, it had to be set on the basis of forecasts, our forecasts of income tax revenue and the OBR's forecasts of the UK economy, which feed into the block grant adjustment. As you said, convener, we would expect that with the Scottish economy growing, Scottish income tax would be growing, and that is indeed the case. We would also expect that, with the UK economy growing, that would feed into the calculation of the block grant adjustment. We were forecasting a growth of Scottish income tax revenue. The OBR was forecasting a growth of UK income tax revenue, and both of those forecasts went into the calculation of the budget. As Susan emphasised, we are still just looking at what we expect to happen when we find out what the outcomes are and the first income tax out-catern for 2017-18 will appear in July, I think, in 2018-19 and in July next year. We are looking at what seems likely to be the case with those out-turns. It still is the case. Scottish income tax, as we expect, will have grown over this period. UK income tax will have grown over this period, but Scottish income tax has grown at a slightly slower rate than we expected. UK income tax looks to have grown at quite a lot of faster rates than it went into the BGA forecast. The reconciliation comes about not really about the performance of the economies, the UK economy and the Scottish economy. It is really about how accurate the forecasts were. In particular, whether the two—I use the word errors carefully—no forecasts are accurate. If the forecast errors on both sides were the same, there would be no reconciliation. It is not even—do you have big forecast errors? If we had the same forecast errors as the OBR, then there would be no problem. It is when our forecast adjustments are out of line that the reconciliation is needed. It is not really about the performance of the Scottish economy as such. It is about how adjustments in the forecasts now seem likely to be appropriate. I understand all that, but that does not explain why we are losing money as a result of the forecast income reconciliation. Is it not right, in terms of what the OBR projected in its figures, that much of that additional growth that was unexpected and higher than it thought it was going to be was a result of increased earnings in the rest of the UK? Yes, you are absolutely right. The single biggest factor in the size of the reconciliation is that UK tax revenue has risen faster than had been anticipated when the budget was set. That is the most important comparison. Faster growth of UK tax revenue than forecast, whereas Scottish tax revenue has grown slightly slower than forecast, but the gap there is not very big. It is the growth of UK tax revenue that has given rise to the big adjustment. The block grant adjustment is bigger than was forecast two years ago. In our budget report in January, we began to explore some of this, and there were questions from myself, from Adam, and from Murdo, who unfortunately cannot be with us today because he is away at another committee. The report that we produced suggested that there was evidence that the differences in income tax growth may be due to that disproportionate level of higher tax payers in the rest of the UK compared to Scotland. If that is the case, to me, does that not suggest that there is a flaw in the way that the fiscal framework operates? We need to address that as part of the review of the fiscal framework when that comes in 2021. Otherwise, we are going to be in a competition here that we cannot win. In the end, it is not for the fiscal commission to make judgments about how the fiscal framework should be changed. That is a matter for the parliaments and the Governments. You are quite right, convener. The lightliest single explanation of the issue is that, in the rest of the UK, there is a higher concentration of high-rate taxpayers, and the recent growth of UK income tax revenue has been concentrated in high-rate taxpayers. It is not just the distribution, but the fact that the growth of UK tax revenue has disproportionately seemed to have come from that group. In our forecast evaluation report in September, we will be looking at this more closely. I would not say that that is a fundamental flaw because, had we and the OBR anticipated that two years ago in setting our forecasts, we would have said that, in forecasting the growth of Scottish income tax revenue, we have to take account of the distribution of Scottish taxpayers. In forecasting the growth of UK income tax revenue, which will go into the block grant adjustment, the OBR will take account of the distribution of UK taxpayers. None of us were aware that the growth of income tax revenue in that two-year period would be so strongly affected by distributional issues. With the benefit of hindsight, we would have done different forecasts. Does that mean that there is a fundamental flaw? No. Had we known about this, the budget would have been set on that basis and we then would not have reconcilated. Scotland would have had a smaller budget two years ago and we would not need a reconciliation now. First of all, it is not a flaw in what you have been doing. No, it is not necessarily—having said, it is not for us to pronounce on the fiscal framework, I should not say—it is not a flaw in the fiscal framework. There is no particular reason to think of it as a flaw in the fiscal framework. Those sorts of adjustments are the normal business of forecasting. Thinking about the distribution of high-rate taxpayers in Scotland versus the UK is a twist that I do not think that either the OBR or us would feel that that was a terrible mistake to have made. It is a twist that, with the benefit of hindsight, we might have done a slightly better job. However, that is the normal business of forecasting. Two years on, with any set of forecasts, you will find things that you could have done better. The scale of the corrections here is the most important message from this. The scale of the forecast corrections is not out of line with what you expect to arise from putting together the forecasts of two forecasters doing a pretty good job of forecasting. Nobody is going to get it right. The message is that the Scottish Government has got to be ready to deal with reconciliations on the scale—hopefully not £600 million every year, hopefully not always negative—there will be some years in which it is positive and negative. However, the scale of adjustment is not out of line with what you would expect with the framework that is set up. I am sorry that I have talked too much, but it is complicated. With all that being said, it still means that, in terms of the paper that you produced, a £229 million problem for the Scottish budget, which might not be much in forecasting in terms of that, but it is a lot of money at the end of the day. We really need to get underneath why that is happening. If there is a structural problem that exists in the UK in salary terms and earnings between what happens in Scotland with a number of additional and higher-rate taxpayers that we have compared with the rest of the UK, that might not be a one-off. Except that, if that is the issue, it is something that we will feed into our forecasts in the future. Forecast errors do not reproduce themselves because the forecasters learn, and we would take that into account in our future forecasts. I think that it might be helpful to separate those two things out. What Scotland will get through this process depends on the structural difference between the rest of the UK and Scotland. You are absolutely right. However, the reconciliation issue is not that. The reconciliation issue, as Alistair says, is about comparative forecast accuracy. You are right that there is a structural issue, perhaps, between the labour markets in Scotland and the rest of the UK. In the end, it will wash out and that will affect how much Scotland gets, but it does not at all influence this reconciliation issue at all. That is to do with forecasting. If we take the baseline and where we started, that means that, from now on in, we are going to be £229 million out every year from now on in because of where we started from. We are going to be £229 million down in 2017-18 looking at the forecast for 2017. That is structural. How is that going to change? The Government will be setting the budget for 2021. In setting the 2021 budget, I imagine that we will all be aware of the implications of the distribution issue for our income tax forecasts and for the BGA forecasts. Those factors will feed into the forecasts. That may mean that the 2020-21 budget is less favourable than it would have been without this, but it is no longer a reconciliation issue. If we get the forecast right, it is in the budget. I do not know what is to come on in this one. I want to just burrow down a bit further, because when we looked at the HMRC's forecast for the number of additional and higher rate tax payers, the forecast that you built your own modelling on was 18,000 additional rate tax payers. The actual number turned out, once we got the outturn figure from HMRC, to be 13,300. If those 18,000 who were in the system to begin with would have a 3 per cent increase in their wages, that is a heck of a lot more money than 13,300 additional rate tax payers having an increase in their wages of 3 per cent. Inevitably, built in from the beginning, we have a problem here. As the minister says, the way that it manifests itself will be shown up in our forecasts in the future. I think that it is really important to separate those two things out. There is a structural issue here. You are absolutely right, but it is not manifested as a reconciliation issue. The structural issue that you need to worry about in a sense is to do with the whole issue of devolution of income tax, which is not our bag if you are a bag. That is where the structural issue that you are thinking about comes in terms of how the reconciliation operates. It is about forecasting and relative forecasting accuracy. Both the OBR forecasts, which will determine the block grant adjustment, and our forecasts take account of the outturn data. They are both adjusted downwards by £550 million to reflect that. There is no impact of that outturn data and the reduced number of income tax payers on reconciliation. There is no net effect at all, but you are right. There is a structural issue to do with the devolution of income tax, but that is not the fiscal commissions. Can I add a footnote to that? As Alistair said, the reconciliation numbers will likely go up and down. They will certainly vary from year to year. You have a budget that is real, based on forecast numbers, which will change and evolve. The Government will need to think, as I said in my comments, about how we manage our budgets at home. How do we allow for variability so that it can pay out in real-term times benefits or make expenditure when those numbers will vary over time? That is another challenge. I recognise that I have been dominating a fair bit of this discussion, and I would like to go further, but I look to fear tellers. I have got to let them in. Thank you, convener. Good morning. I want to ask two sets of follow-up questions in the same area that the convener has been exploring with you. The first is about earnings, and the second is about borrowing, which Professor Smith just said. Earnings first. Last time you were in front of us five months ago or so, you were quizzed by Alexander Burnett and others quite heavily about some very eye-catching numbers that you had then forecast in terms of the projected increase in Scotland of additional and higher-rate taxpayers, which you said would lead to very significantly increased income tax revenues accruing to the Scottish Government in the forecast period from now looking forward for five years. Do you stand by those forecasts because they caught a number of us as being, I think, quite surprising last time you were in front of us? Now what you are telling us is that the opposite seems to be the case and that income tax revenues are not likely to increase in Scotland as quickly as had previously been forecast because of some differential between the income distribution in Scotland compared with the rest of the UK? I think that in December what we said was that on the basis of the out-turn data we had adjusted our view of where people sat in the distribution. We said that there was an increase in the number of higher-rate taxpayers. We also, I think, said that because where they were positioned in that distribution, they were positioned very near the threshold that it did not lead to such a proportionate and dramatic increase in income tax revenue. The paradox that we were trying to explain last time we appeared was an increase in the number of higher-rate taxpayers, which was very dramatic over time, but that income tax revenues did not increase by so much. They did increase, but I understand that. Do you stand by those forecasts? Is it still your view as a Scottish Fiscal Commission that the number of additional and higher-rate income taxpayers in Scotland is likely to grow as you just described it, Mr Ireland, dramatically over that forecast period? The numbers of higher-rate taxpayers, which we have published, we have not put them in our summary report, but we do publish them on the web. Those numbers are consistent with the numbers that we used in December, yes. Do you stand by those forecasts? Can you just remind the committee what those forecasts are based on? Forgive me, but I think that there are a number of us who have a degree of scepticism about the amount that those numbers can really be trusted. The way in which the income tax forecast is constructed is that we have a sample survey of the individual tax records throughout the UK with geographical sort of identifiers. We can identify the Scottish taxpayers. This is a large sample of the initiative data. What we have done with that sample is that we have taken the out-turn data that we had for 2016-17, and that gave us the number of taxpayers in each band. We adjusted the number of people, the number of those tax records, so that they reflected exactly the distribution that we saw in the out-turn data. In other words, you can be confident that we are reflecting the out-turn data, which is the best data that we have in the micro data that we use to do our forecasts, so that we can be confident about that. Then what we do is we apply our macroeconomic forecast, the growth rates from that to the individual tax records, and then we can see where people lie over time in the income tax distribution, where the distribution between tax bands, because that is a purely mechanical thing. That is in essence how the forecasts are generated. This time around, we have had another sample taken of the administrative records, so we have another year's worth of data. We have again adjusted that so that it reflects exactly the out-turn data that we have for 2016-17 in terms of numbers. That is why, in a sense, we stand by forecasts. The impact of those changes has been some changes to that new data. It has not been dramatic. The thing that is driving our income tax forecasts this time around, which is, of course, the difference, is the improvement in our macroeconomic forecast. Earnings are doing particularly well at the moment, and that feeds through into our macroeconomic forecast, and that has generated the new income tax forecasts. I hasten to add that the growth in the income tax forecasts this time around is not as dramatic as you might take from the earnings, because we have made an adjustment to account for some UK policies. We have got better information on how UK policies have been affecting Scottish taxpayers from the OBR, and that adjustment for pension auto-enrollment in effect cancels out most of the increase in the income tax forecasts. There have been two net movements. That is helpful to get all of that on the record. I appreciate that. The second set of questions I wanted to follow up with you on the basis of what the convener was asking. Professor Smith, if I got this downright, I think that you said in answer to one of the convener's questions a few moments ago that the Scottish Government has to be ready to deal with this scale of adjustment. We are talking about £1 billion over the course of three fiscal years of income tax reconciliation adjustment, which you say is not out of line with or not out of kilter with the ordinary business of comparative forecast inaccuracy. The key point that you are making there is that the Scottish Government needs to manage the nation's finances such that it is ready to deal with this scale of adjustment. That is going to be a normal feature of fiscal events and financial planning in Scotland under the current fiscal framework. Given that you have said at paragraphs 367 and 368 of your report published last week that the Scottish Government is not going to be able to deal with this level of adjustment through the use of its borrowing powers, because so much of the credit card to use the colloquial has already been maxed out. Does that not suggest that the current stewardship of the nation's finances has not been sufficiently prudent to deal with this ordinary, usual, routine scale of adjustment? First of all, I confirm what I said earlier. This scale of reconciliation is of a general order of magnitude that shouldn't be surprising and that that the Scottish Government will need to plan for. Hopefully, sometimes it will be positive rather than negative. The Government's scope for dealing with those adjustments without having to cut spending or raise taxes comes from its drawing on the Scotland reserve, where it can draw a maximum of £250 million a year and from resource borrowing a maximum of £300 million a year. It is not so much an issue of maxing out the credit card, but it is that the credit cards have got annual limits on them. However much money is in the Scotland reserve, if only 250 can be drawn down any one year, that is the annual limit. If we have to face an adjustment of £600 million in a year's time—that is, at the moment, just an estimate—it is just a matter of simple arithmetic to add 250 and 300 and say that it does not add up to 600. That is why, if things turn out this way, the Government will need to look at adjusting expenditure or raising revenue to deal with it. It won't need to look at that. It will have to do it, won't it? That is what you say at 368 quite graphically. That will mean that the Scottish Government will have to adjust its spending plans or increase taxes. Those are your words. Can I just draw you out on what you mean by that? Just add to what Alistair said before and respond to something that you said. Just on borrowing, there are two sorts of borrowing. There is resource borrowing and capital borrowing. The Government is using its capital borrowing and it will, in the future, max out if it continues the way it is going. It has yet to use its resource borrowing powers. It is resource borrowing that it can use to deal with reconciliations. It has 300 million on the credit card, as Alistair would say, which it can use to deal with reconciliations. It can use 300 million a year. The issue is that it has to repay that borrowing over three to five years, so it cannot run up a big credit card bill. There is scope there for 300 million borrowing a year. Also on the reserves, Alistair said that you can use 250. That is true. You can use 250 of reserves, but those are resource reserves. At the moment, there is only about 146 million in the kitty. Just to modify a little bit, Alistair said that it is more optimistic perhaps on borrowing than you might have thought. It is slightly less optimistic than the 250, because there is only 146 at the moment in reserves. That is very helpful, but, to underscore a point, you say—again, those are your words—3.67 of the report, that the figures combined will not be sufficient to manage the income tax reconciliations that you are forecasting. That will mean that the Scottish Government will have to adjust its spending plans or increase taxes. Is that right? I think that we say would, because those reconciliations that we have in our report are still forecasts. If we end up with reconciliations of 250, 249 and 608, then the would becomes a will, yes. We have spelled that out in our report, and I do not think that there is anything to add to the words in our report. I have a few people here in supplementary, and I think that, Tom, you were wanting in here. Good morning, commissioners. I just want to follow up and clarify my understanding, and the reality is that there are two credit cards. There is a capital credit card, which has been used, but then there is the resource credit card. The Government has not even taken a ballpoint pen to go and sign the back of it yet. It has not been used at all, so that resource capacity is still there to be used. The resource capacity is there to be used, and they have not used all the capital, but they are using it at a pretty quick rate yet. The element of that is that it is up to £600 million a year, £500 million for in-cash management. That is what would be deployed to manage the demand-driven expenditure, such as social security. Is that correct? The in-cash, the £500 million a year for in-cash management within the resource capacity for borrowing, would be used to meet demand-driven spending, such as social security. The reason that I ask that question is that you make forecasts upon social security, and there is an expectation next year of £3.5 billion. Is that correct? On the social security, yes. Because the social security budget will increase and it is demand-led, there will be a need for in-year cash management, and there are facilities to do that. Can you clarify your confidence in the forecast of £3.5 billion next year, because I appreciate that it is very difficult because these powers are only coming online? I think that it is important to be exactly clear what that represents. Our forecast of the social security benefits that Scotland will have the Executive Authority for, in other words, DWP will minister them on DWP rules, but they are not Scottish rules. It is a broad brush first attempt to forecast what the bill will be on continuing UK rules. If there are changes in eligibility or changes in how the benefits are administered as they are taken over by social security Scotland, of course that forecast will change. We do not take account of that. I want to get a sense of how confident you are. Given the best start grant, there was a forecast, I believe, of 4,000 for the full year, but there were 4,000 applicants on the first day. What I am trying to understand is how confident the commission is in that forecast, so that I can understand whether the in-year cash management resource borrowing powers are sufficient. Are you able to say how confident you are in that £3.5 billion figure? Are we going to, because with all this, there is a very significant forecast, reconciliations over income tax, I am keen to get a sense of what the potential error is in social security forecasting? We would not put any sort of quantitative error bans around that. I think the words that I am using in saying this is an initial forecast based on UK rules and forecasting new social security benefits, as has been shown in the sense that the first one that came along is difficult. I would expect there to be some significant forecast error there. I cannot say which way, but it is a pretty uncertain forecast, I think, is how I would put it. Can you quantify a significant as a potential percentage? No, I do not think that we are in a position to do that, but we are. It is one of the highlights of our report that we think that a social security budget of £3.5 billion with the possibility of eligibility changes coming into that as it gets fully devolved is a sum of money with a high degree of risk. That is one of the highlights of our report. We are drawing attention to the fact that there is a significant fiscal risk in this, so we are agreeing with you on that. Putting numbers on it is harder. You could do—I am not going to do it here on the back of an envelope, on the pocket calculator—what you could do is you could look at the OBR's forecasting record for social security benefits and apply that average error that the OBR has made to our £3.5 billion forecast. You could get a ballpark figure, but we have not done that. As I said, I am not going to do it this morning. Okay, I will not push any further on that. With regard to income tax forecasts, I appreciate that this is an art as much as a science and it is developing in its silver early days. What are your expectations that we could be seeing in the future forecast errors? I appreciate that there are two parties at play here, both the OBR and the OBR. What is the possibility of further reconciliations in the region of £600 million being required? Is that something that we should become accustomed to and expect, or should in general be expect, as the US progress and more information and data becomes available, that we would be expecting far less margins of error? We have had a look at the OBR's track record in forecasting income tax. Susan mentioned at the start that about 3.3 per cent is the average absolute error, so you can expect that magnitude, which roughly does come to about £500 million. We also saw that in the OBR's track record that you get is a pattern of consistent errors. You get a positive error, a positive error, a positive error and then perhaps it will flip. On average, you would expect the errors to be zero because forecasters do not make systematic errors, they just make errors that are then corrected. However, what we observed in the OBR's forecasting record was consistently an error of one sign, an error of the same sign, an error of the same sign, so that they persist over time. What we are forecasting here is three negative reconciliations. I am not going to promise, but I would hope that at some point that flips into the other way round. Just to add to the complications of what John has said, the reconciliations that we have been explaining arise not from one set of forecast errors, but from two sets of forecast errors, which complicates the issue further. It is not what forecast errors might we reasonably expect from the Fiscal Commission, what forecast errors should we reasonably expect from the OBR. However, how often will we err in the same direction so that there is no reconciliation problem? How often will we err in different directions so that our two errors put together create a bigger reconciliation rather than a smaller one? Without really wanting to muddy the waters or the thought process, there will be other reconciliations. It is not just about income tax, so over time we will see reconciliations in relation to social security and some of our devolved taxes, so it is a very complicated landscape. I appreciate that it can go either way, and in the longer run it should effectively cancel each other out. However, £550 million is not an unrealistic number. I think that when we answered the question in a slightly different way, when we saw it was £500 million, £600 million, we weren't surprised. The final point that I wish to make is that the theoretical maximum that the Scottish Government would have to address is £550 million. That has taken £300 million of resource and £250 million of reserve, but that is contingent upon having enough reserve. The resource is also contingent upon, as with a £600 million cap of which £500 million can be spent for in-year cash management. If money is getting used to address social security forecast errors in year, my understanding is that that would be the amount of money that is from resource borrowing available to address income tax forecast errors. The potential of having £550 million to address an income tax forecast error is fairly limited in itself. That was my reason for my line of question. I appreciate that you cannot comment on the applicability to the fiscal framework, but it clearly raises a lot of questions about going into the fiscal framework review. If we can be in a situation where £550 million of forecast error is not going to be abnormal or unusual, then that exceeds what is available for the Scottish Government to meet that. I think that because we cannot comment and it is not our job to comment on the fiscal framework, we took the approach that Mr Mackay took in his statement. Here are the rules and what are the implications of the rules. The implication of the rules is that if those reconciliations turn out to be as you say they may be, then there is a need to adjust the spending or more taxes. As a matter of fact, we can agree with you that the Scottish Government's borrowing powers are set out in our report, a relatively modest set against the scale of possible reconciliations. Just as factual matter, that is a fair update. That is independent of policy and Government decisions. That is just based upon how much information can be provided through forecasting. If those borrowing powers are not enough necessarily to meet forecast error, is that fair to say, just as a statement of fact? I would just stick with the statement of fact that I gave you a moment ago, rather than trying to elaborate. Thank you. Good morning. I apologise for sticking with this area a bit further. The scale of the numbers that we are talking about has been described as normal, but it seems to me that a normal Government would face uncertainty from one set of income tax forecasts rather than two conducted separately, perhaps under separate methodologies, which potentially compound each other. Normal is perhaps a questionable word to use. Just in terms of the scale of the 2018-19 tax year, which is the biggest element of what we are talking about, I think that we all appreciate that we would not expect you to comment on the policy choices that a Scottish Government would face if that is the figure, the 600 million plus that we are talking about. However, Susan Rice still described that as an estimate. To what extent is that based on actual tax data for the 2018-19 year, as opposed to still being an estimate? If it is still an estimate, when will we know what the final figure is? You will appreciate that, not for you to comment on, but for us to consider, there are huge political implications from a reconciliation of that scale on a budget that is being voted on in this Parliament two months before a Scottish general election. The simple answer is that you will not know that particular year's outturn this July, but you will know it next July. That is the straightforward answer. Obviously, the information that we have in the numbers that we will see this July will help us to refine the estimates. Also, the real-time information that HMLC collects and we see about pay-as-you-earn receipts at the Scottish level also feeds into our forecasts. There is some real data, if you wish, in our forecasts already. Over time, that will improve, and the next big improvement will be in July when we get the next outturn data. However, you will not see the outturn data for this particular year until the July afterwards. You say that with the data that we have at the moment, the data that you are working on at the moment, you are narrowing down the amount of uncertainty in that, or should we still consider it a forecast rather than— Oh, it is definitely a forecast. It is definitely a forecast, but as we get more data and as we get RTI data, the real-time data, as we get the outturn data and as a couple of months ago, we got the sample of administrative records, the data element to that forecast is increased and the forecasting element is reduced. It is a gradual process. It is not black and white, but we are certainly getting more information and more hard data that is going to the forecast over time. There will always be, at least as far as we know, the 15-month lag between the end of the fiscal year and the publication of the outturn data for income tax. The hard number will finally be known in the middle of the summer recess a couple of months before the Scottish Government has to publish a budget for 2021-22. That should be noted. Thank you, convener. I think that what is making the committee nervous is the way the forecasts are changing and the fact that we are in a position where the set of forecasts paint a more bleak position than previous forecasts. I am keen to take one year and to get an understanding of that year in terms of how the numbers have moved. In terms of this budget year 2020, from the position where the budget was set, we are now, potentially, based on these forecasts, £188 million worth off. The reason for that is that the forecast revenues that you have set for the Scottish Government tax revenue have increased by £20 million. However, that has been offset by the OBR forecast of the block grant adjustment, which has set an increase of £208 million. I am just interested in those particular movements. First of all, in terms of the forecast increase in Scottish Government income tax revenue, what kind of drove that change from earlier in the year? I think that you are looking at table 4.4, which gives the change in our income tax forecast. As I said earlier when I was responding to Mr Tomkins' question, the improvement in earnings caused us to increase our forecast by £150 million in that year. The big negative of £127 million in our forecast, though those two figures offset each other, was due to us having more information about UK policies and their effect on Scotland. After talking to the OBR and HMRC, who are responsible for estimating the effect of those UK policies, we took £127 million off our forecast to reflect how pension ortoenrolment was working. There were other UK policy changes that added £10 million to the forecast. We also had the new micro data that I was talking about earlier, which was not £30 million off the data. In all, that movement of £20 million, that increase in £20 million in the forecast, was primarily driven by an improvement in our economic forecast driving it up and then a reduction due to our taking better account of UK policy in Scotland. On pension ortoenrolment, it was known back when the budget was set that it was going to come into play. What information came to light that resulted in new factor in our reduction? What happens here is that when the policy is first set, the OBR publishes a set of policy costings. One of the problems with this set of policy costings is that, when their policy costings changes, it is not reflected in the publication. It never revised the publication. It is very hard for us to track exactly how HMRC costings for the OBR change. What we rely upon is conversations with the OBR. The conversations that we had with the OBR before Christmas indicated that there was some change happening, and we tried to take account of it then. However, the subsequent conversations that we had after Christmas made it clear that they were again changing their costing, and we needed to change ours, too. As we spoke to the OBR, we got more insight into how HMRC costings for the UK were changing, and we took those on board. As soon as we had the information, we included it in our forecast. That is how it works. We changed our forecast to reflect UK policy a great deal. It just happens that this was a particularly big one, but it is the same process. We continually taken the public information initially and then conversations with the OBR as they tell us how those things have moved. In terms of the OBR's side of it, that increase in the forecast of the Blanc Grant adjustment position of it began up by £208 million. What was the OBR's basis for that? That was a very different issue. That was the issue that we were talking about at the opening of the meeting. What was happening, in essence, is when the OBR had made their original forecast, they had quite a low growth rate for UK tax revenue, and as they saw how UK tax rates were increasing over time, in part due to the higher earners in the rest of the UK, they revised their growth rates upwards. The movement in the Blanc Grant adjustment was dealt with. That came through a year or so, or two years ago. It was a much bigger adjustment than that. It is not related to the policy costings at all. That £208 million figure has changed from when the budget was set, so what new information has come to light for the OBR to change that in quite a short period of time? Just the continuation of what we have seen already with UK tax revenues have been surprisingly buoyant and the surprises are still coming. That adjustment of £208 million reflects continued growth in UK tax revenue, probably, as we were talking about earlier, concentrated at the top end of the taxpayer distribution. Just before I come to Willie, I just wanted to again try to bottom some of that out. If we would agree, does that structural issue in there? In your evidence, it is now clear that there is a risk that the UK tax base is growing more strongly than the Scottish tax base for 2017-18 and for 2018-19. That might be likely to continue. If that is the case, what is the main drivers in that difference? It is worth emphasising that all those tax adjustments with income tax devolution start off with a base in 2016-17, so that the block grant adjustment for 2016-17 is set equal to Scottish income tax revenue. When we had the out turn of 2016-17, that adjusted for the fact that there were fewer high-rate taxpayers in Scotland than proportionally in the rest of the UK. What we have seen in the past two years, apparently, is that income growth has been stronger at the top end of the income distribution. Whether that is a permanent feature of the economy, we do not know. That is just something that has happened in the last two years. It might well reverse. There is no law. It seems to be a feature of the structure of the Scottish economy that Scotland has fewer high-rate taxpayers, but whether the income at the top end of the income distribution is going to grow faster than the rest of the national income, there is no reason to suppose that that is going to be a permanent feature. Why is that? If you think of the kinds of people who are just into the high-rate taxpayer, it is the highest-paying taxpayer, why have the income of chief executives of commercial lawyers, partners in international accountancy films, and why have their incomes been rising faster in the past two years than the incomes of the rest of people? That is a big question that we could spend hours speculating about. The important thing is that there is no particular reason to suppose that that is going to be a phenomenon that is going to continue. Can I just add an underpin as well to the differences in growth and income tax? We have mentioned these things in the past, which is that the working-age population in Scotland is not growing at the same rate as we see in the rest of the UK. The overall population is not growing here to the same extent. Those may be smaller effects, but that affects how much income is being drawn because our income tax is based on our natural earnings income here in Scotland. It is those demographic features as well. I will come to Alexander soon on that, but I think that Willie still has a question in the area that we are in. If we do not know whether those factors in terms of salary at a higher level are going to continue or not into the future, we better start finding out. Otherwise, this is going to have a potential dramatic effect for some time and the difference in the rest of the UK economy compared with the Scottish economy. My question is really about forecasting in general, but I wonder if you could clarify for me something that has already been said for my simple mind, if you do not mind. He said that the biggest influencer, the biggest factor on what might impact the Scottish budget is not the performance of the Scottish economy, it is not the relative performance and tax take between Scotland and the UK. It is about the comparative forecast accuracy, I think that that is the most John used. Is that the main factor that leads to those reconciliation figures? No, that is not what I meant to say. The reconciliation issue arises because of two forecasts being out of line in a different way. That is where the reconciliation problem comes away. I agree with the convener that if the underlying issue is an issue about income growing in different parts of the distribution at different rates, then once the forecasters get a grip on that, it is built into the forecast and it stops showing up as a reconciliation issue, but it does not mean that the issue has gone away. It would then be built into our forecast and if it is indeed the case that Scottish income tax revenue, for structural reasons, is going to grow more slowly than UK income tax revenue, then that is an objective problem for the Scottish budget. It is showing up as a reconciliation issue at the moment because the forecasters did not anticipate that. Once we start anticipating it, it will be built into the budget and into the forecast. It does not mean that it goes away, it just pops up in the budget instead of the reconciliation issue. It was actually John that said that phrase, because I have written it down and you said that John Tyarski specifically about it. You have just said there, Alasdair, that if there is a relative better performance in the UK for tax revenues that they take, then the Scottish budget gets hammered. That is the way that the block grant adjustment works. No matter what Scotland does in terms of its performance in economic performance, if it is outpaced by performance in the rest of the UK, then our budget takes hammering. That is the way that the fiscal framework works. The block grant adjustment looks at the growth of UK tax revenue after 2016-17 and the growth of Scottish income tax revenue after 2016-17. If UK income tax revenue grows faster than Scottish income tax revenue, that is a problem for the Scottish Government. That is what devolution is about. Devolution means that Scottish income tax revenue depends on the performance of the Scottish economy, not on Scotland getting a share of UK income tax revenue. More than that, is it not? It is about a competition between the rest of the UK and Scotland, not just about how the Scottish economy is doing. There is the fundamental problem. I see that Professor Smith is not responding to it because it is a political point. Alexander, I apologise while I was going to move on. It was leading me into the black art of forecasting. It might be fair to say that you guys are on the pessimistic side of the forecasting business. I do not want to start a spat with Tony Mackay's paper that I am sure you have read. He says that your estimates are always on the pessimistic side. He looked at the GDP forecast, for example, which he says that since December to now, you have downgraded your GDP forecast by a third. In December, you were forecasting 1.2 per cent, and now it is 0.8 per cent. What is responsible in your view for that quite marked downgrading in GDP forecast? How reliable are your estimates? Can we have any confidence that we are accurate going forward? There are a number of issues around what Professor Mackay says, but I will lay out a couple of them. The first thing is that we have made four forecasts. That is not a lot of forecasts, certainly not as many as Professor Mackay has made. We do not have much outside data yet. What we also know is that, particularly with national accounts, they are revised for some time. It is fair to say that the jury is still out on our forecasting record. Every year, in September, we look at our forecasting record and produce a comprehensive account of how our forecasts were not in line with the out-turn data, as we thought it was at that time last September. We will do the same in September. We take forecast evaluation seriously. I do not think that we have made enough forecasts or had enough real data back yet to judge whether we are pessimistic or optimistic, or judge our forecasting record. It is also turning to the forecasts that we have just made and pessimism. I think that it is probably useful to note that both Pricewater has Cooper and the phrase of Arunda also revised their forecasts for 2019 down, both by 0.3 per cent. Everyone is revising their forecasts down. You asked why our forecasts have been revised down. I shall now give you a bit of an explanation of that. The last two years have been a particularly strong couple of years for the Scottish economy. That is main driven by very strong net trade performance. In part, that is driven by the depreciation of sterling. We have two good years for the Scottish economy, when the Scottish economy is operating above capacity. The economy cannot operate above capacity forever, so you would anticipate that growth will come down. It is certainly something that we anticipate it does. The other thing that is happening, as you know, since the OBR issued its forecast about three months ago, Brexit has moved on. There is an awful lot more uncertainty about Brexit than there was. We have more data on business investment, and that is looking weaker. We also know that consumption is relatively soft. We know that Brexit uncertainty is having a strong impact over the next couple of years. The other thing that is going on is around net trade. I said before that the strong performance in the past two years has been particularly down to the very strong net trade performance. It looks like the prospects for net trade over the next two years are much weaker. In part, that is Brexit. In part, though, it is the world economy. I think that the big international forecasting organisations are forecasting that the world economy will be weaker. I think that the IMF produced a forecast yesterday. We certainly have OECD, a World Bank forecast, which is showing similar sorts of things. The world economy is less strong. You see the potential trade wars that the US are embarking upon, between China and the US, and potentially the EU and the US. The prospects for the world economy are not as good as they were. Also, the weakness of sterling is now built in. I think that the reasons for our downward division are Brexit and net trade. I know that you want to come in Brexit, but I said to Alexander that I would let him in next. I have done that twice still, so I better let him in at this stage. I think that we all understand that the forecast has not made pretty reading. There has been a lot of casting around causes and a number of people pointing at population and more specifically Brexit. Population is obviously a combination of birth rates, mortality rates, people leaving and in with migration. The net UK in with migration numbers remain steady, even though the EU national is being replaced by non-EU. The question is that, even if we stayed in the EU single market, would this completely close the growth gap between the UK and Scotland? I am not sure that anyone is really an expert on population issues. We do think that the sensitivity of migration to Brexit issues is a key element because quite a lot of the difference between the Scottish economy and the UK economy arises from the fact that the Scottish population, especially the Scottish working age population, is growing relatively slowly compared to the rest of the UK. That is inevitably a big uncertainty and is one of the main channels through which we would expect Brexit, however it turns out, to affect the economy. If the Brexit outcome is one that makes migration into Scotland relatively unattractive or difficult, then that will have quite significant economic impact. If you are saying that the solution is to increase our population through inward migration, is not the logic of that, we are just going to further damage our productivity numbers, which you are already forecasting as being slow in growth, and that is not the real issue here. We are simply not creating the right skills base here at home. Obviously, creating the right skills base at home is a good thing, and one that any economy wants to keep its eye on. Generally speaking, inward migration from the rest of the EU has been positive for productivity and positive for the skills base, because we in Scotland and in the UK in general have benefited from inward migration of a relatively skilled Labour force. That is not a recipe for saying that we do not need to worry about the skills of native born Scots. Of course we do. Silly, we have had politicians saying that we should be attracting high-skilled workers from around the world, specifically GPs. Are you aware of any countries having such a surplus of such people to satisfy our own failings of creating those skills here? That is an element of speculation that takes us too far outside the scope of the Fiscal Commission. Lady Rice in her opening remarks said that the Fiscal Commission's Brexit assumptions are, I quote, a reasonable basis for forecasts. Why does the panel consider the Brexit assumptions to be reasonable? In general, I will turn again to Alistair, who has a particular focus on those matters. We look at a number of different factors. We do not just say that it is not a binary this or that. We look at a number of different factors, but we start with the base of saying that there will be some sort of orderly exit from Europe. We do not have evidence at this point to say that there will not be an orderly exit, but we do look at a number of different factors that draw out of that orderly exit. It is not just A or B. There are a number of different implications. Even with the uncertainty that we have and uncertainty that has increased, we feel that that starting point in our forecasts is still the right one at this point in time, or the point in time when the forecast was completed. I do, but I would like to hear why you think that the Brexit assumptions are reasonable as we do. Are previous forecasts based on the assumption that there will be an orderly exit in March of this year? That was a reasonable assumption at the time that the UK Government was aiming for. It is what was by and large expected. It did not happen. Anybody looking at the UK scene now would say that the probability of an orderly exit has gone down because there are now a number of politicians vying for an important position in the UK Government who seem to be willing to contemplate a no-deal Brexit. Any reasonable patient would say that the probability of a no-deal Brexit has gone up. Our judgment still would be that that is not the likely outcome. There is not a majority in the UK Parliament for a no-deal Brexit. Our central assumption remains an orderly exit at some point, but, as Susan said, this is a broad-brush estimate. If the out-turn turned out differently, for example, a further extension granted by the EU or a decision to hold a further referendum on Brexit, that would have economic consequences that would fit in our broad-brush assumptions. As Susan said, the one outcome that does not fit within our broad-brush assumptions is a no-deal Brexit, because it would have substantial economic consequences that are not in our forecast, and therefore that is, as Susan said, a substantial risk to our forecast. However, we still think that the range of outcomes that are compatible with our forecast that I have just outlined are taken together more likely than a no-deal Brexit. Given that you have said that the probability of an orderly Brexit has decreased and that a no-deal Brexit or the risk of a no-deal Brexit has increased and a no-deal Brexit is, as you acknowledge, a risk to your forecasts, why, then, in paragraph 39 of your report, when you say that your forecasts are robust to a range of possibilities, what is in your report is based on one scenario? Given that Disney appears to be a consensus in the Westminster Parliament for anything, it is quite difficult to point to hard and fast evidence. I would be interested to know what work the Fiscal Commission has done about various possibilities about various scenarios regarding Brexit. We think that the assumptions in our scenario are broad brush enough so that they, as I said earlier, would be appropriate for any of the non-no-deal Brexit outcomes. If you were to, for example, ask us what would be the consequences of a further delay or what would the consequences be of there being a second referendum whose outcome would be unknown for a bit of time, then that's sufficiently similar to the situation that we're currently in, that at the level that we're looking at of macroeconomic forecasting, it would not make a very substantial difference to us. So we haven't, we've been content with the broad brush approach that has built into our current forecast. What we have done, some thinking about is because clearly a couple of months ago we had to think seriously about would a no-deal Brexit be our central assumption. We decided rightly, I think, as it turned out so far, that that would be not an appropriate assumption to make at that time. We've thought about the issues that would go into a different forecast if we had to make a different forecast based on a no-deal Brexit. So there would be different assumptions about business investment, different assumptions about migration. There would be assumptions about the economic effects in those sectors of the Scottish economy, like agriculture, that would probably be pretty severely affected by a no-deal Brexit. Those sectoral effects would be big enough to have an effect on the overall forecast. So, of course, we've done some preliminary thinking about what would be the issues that would go into an alternative scenario, but we haven't, we didn't and don't produce alternative forecasts, but if the situation changed and there was a new fiscal event at which we had to produce a new forecast based on a no-deal Brexit, then we have done some preliminary thinking about how we would go about doing that. So there are a number of factors that go into our thinking around Brexit. Alistair has just mentioned a couple of productivity for undirect investment, trade, migration, currency and so forth, and we look at those factors. In a few cases, we've done some sensitivity analysis, factor by factor, so that provides a little bit of background and something for people to chew over, as it were, but we don't produce alternative scenarios and alternative forecasts for a situation like that. So, if I've heard you correct, I understand what you've said about another set of work that's all broad brush and that it's all based on a non-no-deal scenario. However, you don't need to be much of a political scientist or an economist or a professor to look around you and see that the political situation is fluid. I'm not asking you to comment on the political situation, but what work around a no-deal scenario or indeed other permeations are you in a position to share bearing in mind that Mr Ireland articulated very well earlier impacts on business investment, trade, migration and productivity? As Susan said, it's not our job to produce alternative scenarios. We base our forecast on what seems to us to be the appropriate scenario in any given forecast. We don't have a no-deal Brexit scenario hidden away, but, as I said, we have done quite a bit of serious thinking about what would go into that. I've already said that among the key features would be effects on migration, effects on investment and effects on particular sectors of the Scottish economy. However, it's also worth thinking about the fact that if we did get into a no-deal scenario, that's a situation in which UK fiscal and monetary policy would probably be adjusted pretty quickly to address the consequences of it. The UK Government would be looking at whether there are appropriate fiscal changes. The Bank of England would be looking at monetary policy. It wouldn't even be sensible for us to jump in and produce forecasts when we are required for fiscal events. It would be sensible in any case to wait until we were clear what the fiscal and monetary framework was in which our no-deal scenario forecast was being done. We might be waiting a while for some clarity. My final question is that, again in your report, you paragraph 37 said that, although there was a period of nearly two years of sustained above-average growth, none the less, because of Brexit, you've had to revise down your outlook as opposed to up. Hadn't it been for Brexit, you'd have been revising up your outlook, that's what it says here. All else equal, this nearly two-year period of sustained above-average growth might have led us to have revised up our outlook for the economy. However, we expect that the on-going uncertainty created by the Brexit negotiation process will limit growth. My final question is that, if it wasn't for Brexit, what would your forecast have been for growth in Scotland? You're inviting me to describe an alternative scenario that we haven't produced and don't produce. For the reasons that John explained earlier, Brexit uncertainty is one of the factors that have led us to downgrade the economy forecast in spite of relatively strong performance over the last two years. However, it's by no means the only one. The most significant Brexit effect is an assumption that Brexit will have a negative effect on investment growth. However, the single biggest effect on bringing the rate of economic growth down, as John talked about earlier, is the assumption that the depreciation of sterling, which has driven strong trade performance for the last two years, has now happened and it's baked in. We cannot expect continued strong trade performance improvement. That's actually a bigger impact on bringing us down from 1.3 to 0.8 than the Brexit effect. The words about Brexit that you just quoted are words about what is part of the negative effect in our relatively subdued forecasts, but it's not the whole story. I think that I've got a couple of supplementaries on the Brexit question. Am I right in saying that? I'll come back to you on the public sector on Emma, and then I think that Murdo's got a supplementary as well. Emma, on you go on the supplementary. I'm picking up from Angela Constance about forecasting based on tighter immigration policies that might be set by the UK. We're already seeing the number of nurses registering and midwives on the NMC reduced by—there's been an increase of people leaving. 4,000 people have left. It's an increase of 29 per cent. We're seeing a fewer people registering on the nursing and midwifery council, and we know the impact of agriculture if we have a tighter immigration regime, whether it's seasonal agricultural fruit pickers or the all-year-round dairy workforce. I'm curious why there isn't a forecast that's much more succinct about the specific immigration impact on Scotland with a tighter immigration regime. Our current forecast does assume that there will be less migration than we have seen in the past. So there is some reduction in migration building, because we already see that in the numbers. As you say, those numbers are not just numbers, we see the effects in particular sectors that have a lot of migrant workers, such as health services, health and social care and agriculture. However, the big effect on migration would come if we were looking at a no-deal Brexit scenario, then I think that there would be sectoral effects that would be so big that we would be thinking through how they feed through to the macroeconomic forecast. So at the moment, some reductions in migration are built into a forecast and we think at an appropriate level. Earlier Susan said that we produced sensitivity analysis of our forecast and in table 2.7 there's actually a set of variants around different migration assumptions. So if you want to build your own forecast, you can have a look at table 2.7 and that gives you a sense of the impact on GDP, employment and average earnings of high and low variants of migration. Murdo's got a supplementary in this area as well. I'll come back to you on your other question. I'll come back to you in Murdo on the other issue that you want to raise as well. Okay. First of all, my apologies to the commissioners. I was at another committee meeting this morning moving amendments to the transport bill, so I missed the earlier part of your evidence. But I had a follow-up to Angela Constance's questions on Brexit and I absolutely understand everything that you said to her about not providing alternative scenarios. I'm going to have a go anyway. Let's just assume hypothetically Brexit gets called off. We're staying in the EU, we're staying in the single market. In that event, would you expect Scottish economic growth over the next three or four years to match UK economic growth or is there more to it than that? Thank you very much for the warning that you're going to ask a question that would tempt us to talk outside a brief. I think that the honest answer to that question is we don't know it's not something that we've thought about. Brexit being cancelled, we think would, to go back to my response to Ms Constance, is not so far away from our broad brush orderly Brexit scenario that we'd regard that as something that would generate very dramatic effects, especially in the short run, because there would be lots of uncertainty effects associated with Brexit being cancelled that might well have negative economic impacts, but the relative performance of Scotland and the rest of the UK in a no Brexit scenario is very hard to speculate about, even if it was our business to do so. I might even be a good news question. I'm interested in public sector pay, and we've heard about higher rate tax payers this morning, but we know that public sector pay has been increasing in Scotland, and obviously most of us, if not all of us, would like to see earnings continue to increase in Scotland across both public and private sectors. For instance, if every additional pound in earnings occurred, what would be the forecast, or what would you expect to be the returning tax revenue on that? Don't construct that sort of redirector. What I can say is that the public sector pay policy of the Government is built into our forecasts. As you say, the generosity of that public sector pay has increased over time, and that's reflected in our forecasts. However, we don't build those sort of redirectors where you can say, if you increase public sector pay, then you get to pan more income tax or 50p more income tax. That's not our job. Last week, I raised a question in chamber about the possible increases in public sector spend as a result of processes that might should have been following the £1 billion spending decision that Westminster made. In response to that, if a barnatised approach might have been placed on that, we in Scotland would have had an increase in £3 billion to our spending. That would mean that we would have an extra £750 million come to Scotland. Would that have been something that we should have been arguing for? You've struck us speechless. That is just hard for us to respond to. That's your call, not our call, for what you should be. It's a feature of the devolved system, the Barnett formula, that the Scottish Block Grant responds to spending decisions on the rest of the UK. However, as Susan said, it's not for us to advocate one thing or another. You said that the biggest factor was on investment, in terms of the Brexit uncertainty that was affecting investment. Is the Brexit uncertainty affecting investment in Scotland more than the rest of the UK? Or is it the same? I don't think it's particularly a Scotland-only effect that is an effect in the UK as a whole. I don't know about the Scotland-UK comparison. I don't know. I do know that the OBR marked down there in business investment forecasts. It's really important to say that the business investment data, both in the UK and Scotland, is really poor. It's shown quite dramatic reductions recently. It's quite hard to judge or to the performance. That's another issue. Last week, the finance secretary delivered to Parliament the medium-term financial statement. He reported to Parliament a number of aspects of what was in your report. However, in his statement to Parliament, he made only very tangential reference to a number of aspects in your report, including an issue that has taken up a lot of time this morning, talking about the income tax reconciliation and the potential £1 billion plus reduction in the money available in the public finances, which subsequently became clear, was a major talking point, as it has been this morning. That information was not made available at that time to Opposition party spokespeople or, indeed, other members who were asking questions of the cabinet secretary, because your report was not published until after the session had taken place in Parliament. It occurred to me at the time—I think that might be reflected in the views of other members too—that having a very partial explanation from the cabinet secretary of the situation didn't allow the fullest opportunity for parliamentary scrutiny. I understand that your data is agreed in the memorandum between the Scottish Government and the Fiscal Commission. I don't expect you to answer that immediately, but I wonder whether you might want to go away and reflect on the timing of release of data to Parliament and whether, in future, it might be helpful if, in advance of a statement like that, Opposition members might have made available to Opposition members, particularly when there are key issues that need to be highlighted, which are not covered in the finance secretary's statement. You are correct that the way that we interact with the Government is laid out in an agreed and public protocol for those interactions, and we have always been very transparent about that. We follow what is in the protocol, but protocols can be looked at and reconsidered, and indeed the first edition of that did change about a year later once we felt our way through it, not around the factors that you are talking about, but just in terms of timings and exchanges. I could say on behalf of the commission that we are happy to at least look at, not making promises about changes, but we would certainly look at the timing of when we lay our report before Parliament, which is what I think you are talking about. We will take that away and consider that. I will add that the protocol is due to be revised this summer, and certainly when we did the last set of revisions, we spoke to Jim about that, and we will also have a conversation with Jim over the summer. I am glad that Murdo asked that question. It is not just Opposition members, it is the whole of Parliament who are considering the information that is available to us. We can ask proper questions of the cabinet secretary. I thank your witnesses to the meeting this morning. It has been an interesting and probably lengthier session that we have had in the past, but a lot of information has been exchanged at this stage. I will now close this public part of the meeting.