Our second item is to take evidence on the recent United Kingdom spending review from Robert Chote, chairman of the Office for Budget Responsibility. I welcome Mr Chote to the committee and invite him to make a short opening statement.
Good morning, convener, and thank you for the invitation to speak to the committee. It is a great pleasure to be here—as always. The last time I joined you to discuss the economic and fiscal outlook was—perhaps fittingly—April fools day last year, following the coalition Government’s final budget. Since then, we have produced two further forecasts: the first alongside the post-election budget in July and the second alongside the spending review in November. To all intents and purposes, you can think of those as two halves of the same fiscal event. The Conservative Government has used them to depart significantly from the provisional tax and spending plans that they had agreed with the Liberal Democrats in coalition last March, setting out their own preferred strategy for the rest of the Parliament and beyond. In talking about what is in the latest forecast, it is perhaps helpful to contrast with the situation in March, so that we pull together the two elements of what the Government has done.
The first point to make is that neither the underlying forecast for the economy, nor the underlying forecast for the public finances, has changed a great deal over that period. Almost all the action of interest has been in the policy decisions that have been taken and the make-up of the remainder of the post-crisis fiscal repair job.
Back in March, we were predicting that the economy was pretty close to full capacity, that it would grow by about 2 to 2.5 per cent a year over the next five years and that inflation would move relatively slowly back to its 2 per cent target. Effectively, we made the same predictions in November and those are broadly in line with the average views of other forecasters. Those are our central forecasts. History suggests that reality will be less smooth than that, but we think that the fluctuations are as likely to be above those numbers as below them.
Most of the key uncertainties and questions around the economic forecast are pretty much the same as they were in March. When are we going to see a return to sustained robust growth in productivity and wages? How is the economy going to rebalance in response to the continuation of fiscal consolidation? How will the UK respond to global events, such as tighter monetary policy in the United States, lower growth in China and developments in Europe?
Reflecting the recent stability of the economic forecast, the changes in the public finances forecast over the past year have also been relatively small compared with those that we have made in earlier years. Following the autumn statement, many people latched on to the famous £27 billion—over the next five years—that we had apparently found down the back of the sofa. The biggest contributors to that aggregate improvement in the budget balance over the forecast period were a fall in the Government’s prospective debt interest payments, the knock-on effects of the recent strength of some tax receipts and some changes to the way in which we forecast VAT and national insurance contributions. Those positive developments were partly offset by the impact of lower share prices on tax revenues, as well as by judgments that we made on the outlook for spending on disability benefits and property transactions.
Unfortunately, £27 billion is not as much as it sounds. Over five years, it corresponds to an average downward revision to the budget deficit of about 0.25 per cent of gross domestic product. That is pretty small beer in an economy where the public sector is spending about 40 per cent of GDP and raising about 36 per cent of GDP in revenue, and where the average error in forecasting the budget deficit at an autumn statement, even over the remainder of the year that you are already in is twice as big, at 0.5 per cent of GDP. If we add in the changes to the forecast that we made in July, which went in the opposite direction, we have an even smaller underlying net improvement in the budget deficit since March, of closer to £10 billion or, cumulatively over five years, of about 0.1 per cent of GDP. Again, that excludes the impact of policy measures.
By way of comparison, if you look at the underlying changes that we have made to our budget deficit forecasts between previous March budgets and autumn statements, we had a deterioration of about 1.5 per cent of GDP in 2011, a deterioration of 1.25 per cent in 2012, an improvement of 0.75 per cent in 2013 and a deterioration of 0.25 per cent in 2014. The lesson from that experience is that what the sofa gives, the sofa can easily take away; the sums lost or gained have often been much larger than they were last November.
Confronted by our relatively modest changes to the underlying economic and fiscal forecast, what policy judgments has the chancellor taken, taking the budget and the autumn statement together? The key decision that he has made has been to reshape the remainder of the fiscal consolidation—the fiscal repair job—to rely less on cuts in public services spending and more on tax increases and welfare cuts than was implied by the coalition’s plans in March.
The tax increases and the welfare cuts build up gradually and less quickly than the chancellor said that he was going to aim for ahead of the election, so he has also decided to borrow more over the next three years to reduce the severest squeeze on public services spending in the middle years of the Parliament. He then aims for a slightly bigger surplus in the medium term. Combined with the changes to the underlying forecast, that leads him on course to achieve his objective of getting the budget back into surplus in 2019-20, with about £10 billion to spare. The performance of past official forecasts—ours and the Treasury’s—suggests that that corresponds to about a 55 per cent chance of achieving a surplus in that year, on current policy. It is by no means a done deal.
For public services, the two-stage loosening of the belt, as it were, in July and November means that the Government is now looking at a real cut of around £10 billion a year by 2019-20, which is much smaller than the £42 billion a year by 2018-19 that had been pencilled in by the coalition. That corresponds to a real cut in public services spending of about 1.1 per cent a year over this Parliament, compared with 1.6 a year over the previous one. That is smaller, but it is still quite a challenge. A lot of low-hanging fruit has already been plucked and the cuts are obviously much bigger than that average number in those areas of spending that are unprotected.
On the welfare spending side, the Government announced a significant package of welfare spending cuts or cuts in benefits and tax credits back in July, the proceeds of which the chancellor banked when setting his welfare cap—the cap on the cash spending on a large subset of welfare spending. However, by November that cap had already been breached, thanks in part to slower than expected progress on disability benefit reform and in part to the chancellor’s decision to reverse the major tax credit cuts that he announced in July. That is costly in the near term, but less so in the longer term, because by then most of the people who will be affected are expected to have moved on to the new universal credit. The universal credit was also reduced significantly in July, but those cuts were not reversed in November.
To wrap up, if you look at the evolution of the forecast since the end of the coalition, the underlying economic and fiscal picture has not changed very much. We expect the economy to grow, but at rates slower than you would expect in a typical economic recovery. We expect the budget to get back in modest surplus over the next five years, with the deficit having now more than halved from its postwar peak. However, as always there are lots of uncertainties in the underlying forecast, not least the outlook for productivity and wage growth and what that means for tax revenues.
It is also important to bear in mind that our forecasts are based on current policy; they are an assessment of the most likely outcome under current policy. Many other forecasters may think that policy might change. Some forecasters would look at the public services cuts and ask whether they can actually be delivered, some would look at the welfare savings and ask whether the Government can deliver the reforms logistically and the cuts politically, and some would say that if there are disappointments on either of those two fronts or on the underlying forecast, that might mean that the Government will rely more on tax increases or will look again at the ambition to be running sustained budget surpluses into the future. There are lots of uncertainties but, fortunately, those policy ones are outside our remit and we can leave them for other people to worry about.
I am happy to take questions on that.
Thank you. As always, that was a fascinating introduction. Your executive summary on the economic and fiscal outlook also makes interesting reading.
I know that colleagues have questions to ask, so I will not exploit my position in the chair—they all smirk, but it is true. I will not ask too many initial questions, but I certainly want to touch on a number of areas that I am sure my colleagues will want to delve into in greater depth.
I want to ask about borrowing first. In your introduction, you talked about falling debt interest and payments leading to some of the Government’s recent decisions about the money found down the back of the sofa, as you said. Compared with the July forecast, you have revised net borrowing down £0.6 billion to a still quite eye-watering £73.5 billion. In paragraph 1.4, you say:
“Spending on debt interest is also lower in all years, reflecting a further fall in market interest rates.”
What would be the impact of a rise? There has been a rise in interest rates in America. What would be the likely impact on finances if there should be even a modest rise in interest rates?
You can see the sorts of changes between the July forecast and the November forecast. On average, interest rates or yields move by about 0.4 percentage points. If you look at the aggregate for the five years, you see that that contributes about £17 billion of the famous £27 billion.
In July, those numbers moved somewhat in the opposite direction. If you were looking at an uncertainty where money can come in in one forecast and go out in another, that would be an obvious place to look. It is worth bearing it in mind that we simply take the interest rates that are out there in market prices and apply them. We are not making our own subjective judgment of whether the market is over or underegging the level of interest rates.
If you go back a few years, lots of people were saying that interest rates cannot fall that much lower but they have. To date, we have not seen a big movement in the opposite direction in response to the Fed’s move but, obviously, with the new year under way, people are thinking about the outlook for the economy, the expectations for the pace of future interest rate movements in the US and what will flow from that. Things are obviously uncertain and it is one of those areas in which relatively modest movement can make quite a big difference to the fiscal numbers in both directions.
We probably had this conversation a while back and asked whether, given that the interest rate had gone down, it could go down much further. Of course, it has.
Will there not be an impact from, say, a 0.25 per cent increase in the interest rate, or indeed a decrease?
We have a ready reckoner in the outlook. I would dig out the precise link, but I will probably not be able to find it straight away. As I say, a change of about 0.4 percentage points has contributed the change that you can see in the outlook report, so you can scale down to 25 40ths of that.
Fair enough.
In discussing GDP growth at 2.4 per cent, you talk about high population growth, an increase in mortality among older people and the fact that higher net inward migration has driven some of that growth. How much of the growth in the economy is per capita? That gives a look at the real underlying strength of the economy and the extent to which it is just growth from the increasing population.
I can dig the number out. Obviously it depends on whether you are looking at per capita by head of population or working population. The impact of net inward migration increases overall GDP growth as well as the population, but it also increases per capita GDP growth because net inward migrants are more likely to be of working age than the native population on average. It is not the case that such things are simply feeding through into headline GDP for growth but not into per capita GDP.
You say:
“we continue to expect employment growth to slow as productivity growth picks up.”
We have raised the issue of productivity previously and how we boost productivity is a concern. What change in productivity are we seeing at the moment?
09:45
We have seen some improvements, although we have had a set of GDP revisions since we published the forecast and we have not had a chance to incorporate it yet. There have been some downward revisions to GDP that might have affected the overall productivity provision. We have seen some quarters of good news in comparison with what has come in the past, but in the forecast we took the judgment that it was too early to assume that the flowers that we can see are representative of a much better outlook.
We continue to assume that productivity growth is picking up and going back towards historical levels, but we and others have made that prediction for some time and it has taken longer than anticipated to happen. As you know, there is a long list of potential explanations of why productivity growth has been as weak as it has, and one that people place a fair amount of emphasis on is that the difficulties in the financial system have got in the way of the efficient allocation of capital away from relatively unproductive firms and towards more productive ones.
Looking at the conditions in the financial system, I think that it is fair to say that things are easier than they were, so it would be a bit of a surprise and somewhat alarming if we were not seeing some signs of good news in response to that. However, we took the view in the forecast that we should not yet assume that we have kicked straight back to more historically average performance and therefore assume a continued slower return. That matters a good deal in the forecast because we assume that improvements in productivity would underpin an improvement in wage growth, and we need a return to robust real-wage growth in order to bring about the growth in income tax receipts that we are looking for.
You have just mentioned robust real-wage growth, but in paragraphs 1.3, 1.5 and 1.23 of your outlook, you express concern that the new apprenticeship levy will have an impact on wage growth of about 0.7 per cent. You say that by 2019-20,
“an £8.0 billion increase in total departmental spending is largely offset by a £7.2 billion net tax increase (mostly the new apprenticeship levy and larger rises in council tax).”
You go on to say:
“the ongoing costs of autoenrolment and the introduction of an apprenticeship levy will weigh on earnings growth. These are both economically equivalent to payroll taxes, so ... we assume that most of the cost will ultimately be borne by employees.”
You did not really talk about that in your introduction. Will you comment on your thinking on that?
The view that we have taken of the apprenticeship levy is to think of it, in effect, as a payroll tax. Most of the impact will be felt on wages at the end of the process, although there are uncertainties about the impact of measures such as that one, the minimum wage—although that is more on the employment side than the wages side—and auto-enrolment.
Ultimately, the underlying question of whether there is going to be a return to historically average rates of productivity growth and the wage growth that is associated with that is going to be much more important than the uncertainties around those measures, but we have made an adjustment and it is one of the reasons why we have a somewhat weaker picture for earnings growth and income tax towards the end of the forecast.
One of the reasons why the improvement in the underlying fiscal forecast between July and November is at its biggest in the middle of the Parliament and becomes less towards the end is that we have that weaker view of earnings growth further out, hence the Government having to rely rather more on tax increases towards the end of the forecast period in order to achieve what it wants to achieve on the bottom line of the budget deficit. Our forecast is not helping as much in 2020 as it does in 2018, and the judgment on the apprenticeship levy and earnings growth is part of the explanation.
Another issue that you did not touch on in your introduction but which is in your executive summary—at paragraph 1.13—is asset sales. You say:
“asset sales make the difference between debt rising and falling as a share of GDP in 2015-16, with £30 billion expected in the financial year as a whole and £24 billion realised to date.”
You go on to say:
“when the Government gives away some of the assets, as with Royal Mail shares and the planned retail offering of Lloyds shares early next year, the sale will raise less than the asset is worth and the public sector’s net worth is reduced.”
Will you expand on that?
For some while during the last Parliament, we were forecasting that, on existing policy, the Government was not on course to achieve its old fiscal objective of getting the debt to GDP ratio falling in 2015-16, and for some while the Government had basically said, “Okay—we accept the forecast, but we don’t think that it would be sensible to further tighten policy in order to bring that about.”
What then happened is that the Government announced a significant programme of asset sales in 2015-16 that is sufficient to bring down the debt to GDP ratio in that year. In subsequent years, the underlying primary budget balance—the budget balance excluding interest payments—has improved to the extent that it gets the debt to GDP ratio falling of its own accord, such that it is not necessary to rely on asset sales in the same way that it is in 2015-16. We wanted to make it very clear to people that that was why the target had come back into play when it had been out of play for some time.
With asset sales, one of the issues that we always address in the forecast is whether we can be sufficiently confident about what the Government is intending to sell, when it is going to sell it and what price it might get for it to put it in the forecast rather than merely cite it as a potential risk or a potential change. In the scrutiny process that we have gone through, we have been very clear with the Government about what level of certainty we need to have on particular sales for us to be willing to include them. That said, there are some uncertainties. For example, we now expect the student loan sales to be later than we did in the forecast that we produced in July.
In relation to your final point on the impact on net worth, a key issue is that, if the Government is selling something for roughly what it is worth, the underlying health of the public finances is not improved as a result—it is simply getting money today instead of a flow of money in the future. There might also be instances in which the Government is deliberately disposing of something for less than its value. The obvious example is giving some shares to the employees as part of a sale of shares. That will bring about a diminution in net worth, other things being equal, but the Government obviously has other policy objectives in doing that. When those sorts of things happen, we feel that it is important to highlight that. When there are such changes in net worth and an asset sale is made up front that will have an implication in reducing revenue many years into the future, it is worth highlighting that to people as one of the consequences of what the Government has done.
I want to ask you a couple of questions on devolved taxes, after which I will open up the questioning to colleagues. My first question is on land and buildings transaction tax. When one looks at the OBR’s predictions of LBTT revenue, it looks as if there will be huge growth between 2015-16 and 2020-21—it is predicted that the revenue stream will almost double. However, when one looks at what the Scottish Government is predicting, your predictions seem quite moderate. For 2020-21, the difference in revenue appears to be of the order of some £90 million—that includes non-residential and residential transactions.
Could you talk us through how you came to those figures? To me, the predictions of the Scottish Government and the OBR appear to be very ambitious, even taking into account the predicted rise in house prices and the number of transactions over that period.
Sure. The key reason for much more rapid growth—this will be true of our forecast and that of the Scottish Government—is a combination of house price increases and a continued recovery in the rate of transactions that will bring it closer to its long-term average. That combination will provide a much faster increase in revenues than we would see in income tax or VAT—in other words, taxes on labour, income or spending.
The figures that I have for 2020-21 suggest a difference of about £35 million. I am making a comparison with the forecast in the draft budget—in other words, the Scottish Government’s more recent forecast, rather than the one that we use in our report. The Scottish Government’s forecast is about £70 million higher than ours on the residential side and about £36 million lower on the non-residential side. Overall, the Scottish Government’s forecast is about £35 million higher.
On the residential front, the difference is explained by three factors. The Scottish Government is assuming slightly more rapid increases in house prices over the period—about 28 per cent in aggregate between 2015-16 and 2020-21, whereas we have about 26 per cent.
The Scottish Government has a 15 per cent increase in transactions between 2015-16 and 2020-21; we have about 9 per cent. The Scottish Government’s transactions forecast, as I understand it—commission colleagues will correct me if I am wrong—is based on reaching a long-run average of a 6 per cent rate of transactions compared with the size of the housing stock. That is the average recorded between 1995 and 2014.
In our July forecast and earlier forecasts, we were using a similar approach. We assumed that UK transactions reverted to their rate from 1991 to 2004 on average. However, in this most recent forecast, we drew on research that suggested that more of the housing stock is buy to let and buy-to-let properties tend to be bought and sold less frequently than owner-occupied or other properties. We therefore adjusted the forecast down slightly. I suspect that that may be one element in explaining the difference.
The uncertainty around transactions means that the difference between our forecast and the Scottish Government forecast is small compared with the uncertainty that lies around both of them. It is partly because the Scottish Government has taken a slightly more optimistic view of prices and a more optimistic view of transactions.
The modelling approaches that we use are also slightly different. It may be that the Scottish Government uses a technique that we have been anxious about using in the UK context because we have more of a problem with the importance of very high-value properties in London and elsewhere. There may be a difference in the implied number of relatively high-value properties in each of our two forecasts. However, as I say, I would not regard the difference between the two as being large compared with the uncertainty that lies around either of the forecasts in isolation.
Thank you for that comprehensive response. There was so much scurrying about because we were just checking that we were comparing apples with apples rather than apples with oranges in terms of the figures.
It is never entirely clear which year we are looking at and which comparison—
I did say that I would ask a couple of questions but I am actually going to ask three—I have just two more, honestly, folks, so I ask for your patience.
As regards the block grant adjustment mechanism, the draft budget says:
“Until a permanent agreement is reached ... the Scottish and UK Governments have agreed a provisional one-year block grant adjustment for the fully devolved taxes in 2016-17 of £600 million.”
There is no further information in the draft budget on the basis of the figure and how it was arrived at. Is it reasonable or perhaps slightly high or slightly low?
I think that you are taking me beyond my remit. These are deep and treacherous waters, convener.
One can but try.
We simply provide the numbers and it is for others to put them through the political mill and reach policy conclusions from them.
Fair enough—I will not press you on that.
My last question at this point is about landfill tax. You have significantly changed your revenue forecast on landfill tax from July last year and the current financial year from £94 million to £140 million. Indeed, over the next few years, up to 2020-21, the figures that you are forecasting go down from £140 million to £120 million and then back up to £140 million. Those figures are substantially different from the Scottish Government’s figures.
In proportionate terms, the Scottish Government predicts that for 2016-17, the revenue from landfill tax will be £133 million as opposed to your own figure of £131 million. The Scottish Government then predicts that the revenue will go down steadily year on year, from £123 million, to £114 million, to £104 million and then to £94 million in 2020-21.
Why are you at such odds with the Scottish Government on that forecast? Why is there a dip then an increase in predicted landfill tax revenues and why does the divergence seem to grow year on year?
10:00
The uneven pattern is due to the interaction of what we assume about the volume of landfill and movements in inflation over the period. The reason why our numbers are considerably higher than the Scottish Government’s is that the Scottish Government is basing its forecasts on the assumption that its landfill targets will be achieved. Our forecast is based on current policy rather than current policy ambitions. If we had evidence that there were policies in place to deliver those ambitions, we would be inclined to incorporate those sorts of effects.
I note that the Fiscal Commission’s comment on that is that it was
“broadly satisfied that there are potential policies which could feasibly deliver the target which underpins the forecast”.
The key word there is “potential”. There is a methodological difference rather than a disagreement. We would incorporate the impact of policies that had been announced and that we were persuaded were going to deliver, whereas the Government, perfectly reasonably, has produced a forecast on the basis that the objectives will be achieved. The commission says that it thinks that there are policies that could do that, but they have not yet been implemented. If they were implemented, then I would expect the numbers to get closer together.
Thank you very much for that.
I will now open out the session. The first of my colleagues to ask questions will be Mark McDonald, followed by Gavin Brown.
Thank you very much, convener. Good morning, Mr Chote.
I want to delve a little bit more into the £27 billion. I was intrigued when you said that that is not that big a number: I would not say no to it, but I can see the point that you were making. I want to test the vulnerability of that figure.
The convener focused, quite rightly, on the impact of interest and on the impact that an interest rate change would have. We have seen that movement in America and that will obviously set the hares running globally.
What is the vulnerability of the forecasts for tax receipts, VAT and national insurance over the piece? How robust will those be over the period that the £27 billion is realised?
There are considerable uncertainties around most of the forecasts. I will look, in particular, at what you might think of as the two modelling changes that we have made. One is in the VAT forecast, where the problem is that we have underestimated the amount of VAT receipts by a particular route, in that there are deductions from VAT that partly reflect the fact that there are transactions that take place within Government that, at the end of the day, do not generate money for the Exchequer, because you are taking with one hand and giving back with the other.
Probably since before we were around, the forecast models for VAT have assumed that VAT reductions essentially move in line with their historical trend. The problem is that that works okay if public expenditure moves on a relatively straight line. However, if there is a period in which public expenditure falls, then the amount of VAT deductions arising from VAT refunds is lower than expected and therefore the receipts end up being underestimated. That is a long-standing and relatively untransparent model. We identified that that is a mistake in the model that would not become apparent until there was a clear change in the direction of public expenditure. There is a change, so we have corrected that mistake and the model will, I hope, produce a better forecast for the VAT refunds.
We have also been anxious, for some time, about how to model national insurance contributions. For historical reasons, NICs have been forecast using a model produced by the Government Actuary’s Department that is quite untransparent and hard to reconcile with the income tax forecast. There would, for example, be movements in earnings that did not generate the sorts of changes in the NICs forecast that might have been anticipated.
We have said that we will now forecast NICs on the same basis that we forecast income tax, so that there is a comparable model that is more transparent. We should also be able to better check that what is going on in the income tax forecast is consistent with what is going on in the NICs forecast. That has the effect of making that change. It boosts things, probably because there is a more accurate picture of the amount of national insurance contributions being paid above the upper earnings limit on relatively high earnings. We think that that is a bit better.
Those are methodological changes addressing what we think will be sources of on-going error in the rest of the forecast. Of course, there are uncertainties around those.
Obviously, a lot of other things are moving within the £27 billion or within the changes and the underlying deficit forecast, as you mentioned. As the convener said, there is debt interest, which, given how we do the forecast, will simply move in response to movements in market rates. The largest negative element that we have subtracted was the movement in equity prices between July and November 2015. When equity prices are weaker than anticipated, money is lost from taxes such as capital gains tax and inheritance tax. Equity prices may change, so we make a relatively simple assumption about share prices moving in line with the cash size of the economy. That is an obvious source of uncertainty, too.
The Bank of England has suggested that it will start reversing quantitative easing a bit earlier than it otherwise would have done, which is another change. That will produce debt interest savings.
There are also uncertainties around particular sets of receipts coming in stronger or weaker. Most of the big taxes came in a bit stronger than anticipated in November 2015 relative to July 2015. There was a decision about whether that was new news or temporary good news that will go away again, so the issue is about how much that change is pushed through into future years of the forecast.
As I said, you have only to look at the size of the equivalent changes between previous budgets and autumn statements to see that what the sofa gives, the sofa can take away. Obviously, it is for the chancellor to decide, knowing that those uncertainties are there, how much room for manoeuvre he wants to put into his policy plans to achieve the objectives and formal and informal targets that he has set himself.
I noted that, in the OBR report, you have suggested that the “giveaways”, as you have called them, amount to around £18.7 billion. Obviously, that is less than £27 billion, but it is still a substantial amount. Given what you have outlined here, there are a number of uncertainties around the robustness of the £27 billion figure. It strikes me that it would not take a lot to happen for that figure to diminish significantly, depending on the changes that take place over the period. That would lead to concerns that, even if there were a reduction, the reduction could go beyond the apparent cushion between the £18.7 billion and the £27 billion.
You have mentioned a number of areas where delays in policy initiatives, particularly around welfare, led to money that would not have otherwise been made available had those initiatives been followed through. For example, you referred to the slow progress on disability benefit reform. Some people will be quite glad that that has not progressed as quickly as it was projected to. Assuming that progress on that reform were to pick up, would that have an impact on the projections or is that possibility factored in?
I will pick up your first point to begin with. The one thing to bear in mind with the comparison of the aggregate giveaway versus the aggregate improvement in the underlying forecast over the period is that, under normal circumstances, one thinks of an improvement or a deterioration in the forecast getting bigger over time. For example, there may be a policy giveaway or a takeaway that gets bigger over time, so what is going on in year 5 is like what is going on in year 2, only more so.
Both for November 2015 and July 2015 that was not the case. There was no deterioration or improvement that got larger over time; rather, there was a relatively small improvement in the underlying public finances forecast that was focused in the middle years of a session, and which therefore ebbed away towards the end. Equivalently, the policy giveaway was front loaded. In a sense, the Government will give away about £6 billion next year in additional public services and welfare spending by reversing the major tax credit cuts. However, it is no longer doing a giveaway at the end of the forecast, because the public services spending increases are smaller and the welfare giveaway that we had to begin with will have ebbed away because, by then, most people will be on universal credit.
Therefore, the impact of the forecast changes and the policy is more substantive in the middle years of the Parliament than it is towards the end. The chancellor’s objectives for where he wants to be are slightly less vulnerable to all that at the end of a Parliament than in the middle years because there is more going on in the near term than there is towards the end.
I get that, but does that mean that, if there is front loading of spend and if the vulnerabilities were to be realised, there would be a steeper cliff face? If, for example, the £27 billion of revenues that you expect around the middle of the Parliament were not realised for whatever reason—be it that interest rates are moved up or that the vulnerabilities that you mentioned on tax receipts and national insurance contributions materialise—will that make the cliff face a bit steeper because the spend has been front loaded rather than spread across the piece?
It could certainly affect the profile of the borrowing. Some of the uncertainties, such as a change in interest rates or movements in share prices, are probably less likely to have the same lumpy profile in terms of adjusting the forecast and might have more effect at the end than in the middle. It would change the way in which borrowing moves over time and we might find that a positive or negative surprise shows up proportionally more at the end of the forecast than in the middle or in what is visible now from what happened between July and November.
If there are movements in those things, it is for the Government to decide whether and to what degree it wants to change policy to affect the profile of the deficit through the next five years or whether it is really focused on where it is in 2019-20, when it has a target to be in surplus, or thereafter. The overall policy changes that were made between March and July show that the Government has been willing to borrow more in the middle years of the Parliament but is still aiming for a slightly bigger surplus towards the end. It has been willing to let the deficit take the strain in the middle but is not going for a looser position towards the end. It will have to make such judgments every time there are changes in the forecast.
The point that I was making is that, if the bump that was expected as a result of that £27 billion uplift is in any way deflated below the levels of input that the chancellor is making in the front loading and tapering that you mentioned—I would not suggest that it would completely fail to materialise—it could lead to him standing up mid-term and saying that he has to borrow more or cut more deeply to make up for the £27 billion that he expected but that will not materialise.
If the improvements in revenues in the middle period do not materialise to the degree that is anticipated in the forecast, he has a choice about whether he responds to that by finding some revenue from somewhere else through tax increases, cutting spending, which could be in public services or welfare, or borrowing more. Between March and November, he showed himself willing to borrow more in the intervening years to make the toughest of the squeeze on public services less tough when he has not raised sufficient money from tax increases and welfare cuts on their own to do that, so he might do that.
There are different choices depending on whether he changes policy to affect the middle years of the forecast or is focused primarily on the fact that he has a target to get a surplus in 2020 and a revealed preference for having something like a £10 billion surplus in that year. He can respond to those choices in different ways. There is nothing automatic about responding in a particular way if there is a particular weakness in revenues in the middle of the Parliament.
The chancellor has his formal targets, but he also has what he wants to see the rest of the profile of the deficit looking like. There are many ways in which he could respond to that. There is no definite requirement that, if there is a weakness there, he has to go back and revisit the spending plans, for example. They are now pretty much inked in through the remainder of this Parliament.
10:15History suggests that it is not impossible to go back and change those plans, but they are not like a normal budget or an autumn statement. We know that another one of those is coming around the corner and things can be tweaked again in the same way. There are slightly longer-term consequences, as spending plans for departments over that period are inked in, so the stakes are somewhat higher. However, as I said, plenty of responses could be made on the spending side or the receipt side.
The question that I initially asked—before I sidetracked you—was about the initial slow progress on some welfare reforms and whether they will pick up pace or be implemented effectively over the next Parliament. We will leave aside the ideological differences that exist. What impact, if any, will that that have on the projections? Has it been factored in?
One of the reasons why we pushed up the forecast for disability benefit spending is that there is a money-saving reform that is designed to move people to a less expensive system. However, reassessment of people in moving from the initial system to the later system is not going as quickly as expected. There are also changes that are related to the proportion of people who are expected to succeed in receiving the new variant of the benefit. There may be changes in the case load in respect of the number of people who apply. If a money-saving reform is going less quickly than expected, that is obviously an upward pressure on spending. If it ends up moving more quickly than expected and saves the amount that is expected per person in moving from the old regime to the new regime, the deterioration, or forecast, could move in the other direction.
There is a separate set of forecast changes due to the fact that the roll-out of the new universal credit benefit, to which people will move over the longer term, has been pushed further out into the future. In most of the forecast changes that we have done, that does not cost the Government money; rather, it saves it money because, in the original plan, the universal credit system was going to be a more expensive and generous system than the one that it replaced. However, the difference has narrowed as changes have been made to the future generosity of universal credit.
We have been struck that that is following the same pattern as we saw with changes to incapacity benefit. Recent history suggests that such reforms do not move as quickly, and do not save as much money as fast as the Government had hoped. Our forecasts have reflected that, so we have ended up pushing up our spending forecasts. Each time we look at the matter, we try to not get ahead of that, but instead to get the right way and to respond adequately. There remains uncertainty, which we have flagged up.
Finally, I want to follow on from the convener’s questions about LBTT forecasting. Obviously, you have made a significant downward revision of your initial forecast, and that has been done in a staged way. I think that you have reduced your forecast on more than one occasion, and it now seems to be more in line with what the Scottish Government predicted. However, you seemed to indicate that you are not using the same method as the Scottish Government uses to make the forecast. Why was your LBTT forecast much more optimistic initially but is now moving towards being more in line with what the Scottish Government predicts, if you are not using the same method?
The reason for the reduction in the forecast is primarily a story about the UK as a whole, rather than about Scotland relative to the rest of the UK. It comes back in part to the judgment that we made that transactions will not increase as much over the five years as we anticipated originally. We took on the evidence that there is a higher proportion of buy-to-let properties, which are likely to turn over less frequently than they did previously. We assume that that applies in both cases.
The Scottish Government has a stronger forecast for transactions for residential LBTT and a slightly stronger forecast for prices. I suspect that, despite the difference in methods, things are going in the same direction, and I do not think that that methodological difference is statistically significant, by any means. As I said, it is not a disagreement; we do not think that the Scottish Government method is wrong. Its process uses what is called a log-normal distribution—which academic colleagues will be able to explain more correctly and in much greater detail than I can—to fit a distribution, which it then moves. We use microsimulation—a more detailed picture of the distribution of house prices—then move that forward.
The difficulty of using the Scottish Government method in the UK context is that it would not put enough weight on what is going on with relatively high-value properties. For the UK as a whole, what is going on with very high-value properties, particularly in London, is much more important to the forecast. We have seen weakness in stamp duty receipts reflecting a fall in the number of properties that are sold at £2 million plus. There are not many of those, but they account for quite a large chunk of stamp duty receipts. I would not place too much emphasis on differences in methods as a source of major differences in forecasts.
As I said, the difference between the two forecasts now and beforehand is not significant relative to the uncertainty around either of them. The primary source of uncertainty is probably what goes on with transactions in aggregate.
The UK Government will use your forecast as its basis for discussions on block grant adjustment, which is why I wondered about the change. We had a big discussion about the block grant adjustment in the initial stages of LBTT coming into effect. Anything that could prevent such wide disparity in the future would be welcomed.
We have a good relationship with the Scottish Fiscal Commission. We have regular discussions with the Scottish Government and although the Fiscal Commission is not in the room for them, it is there electronically. At those meetings we discuss with HM Revenue & Customs our pre-measures forecast in each round, and we ask whether there is particular information or interpretations of recent outturn data that we should take on board.
I can confirm that we would certainly not make any methodology decisions based on what consequences they might have for a discussion on the block grant. We are producing the best forecasts that we can, based on our professional judgment, and it is for the UK Government and the Scottish Government to use or misuse them as they see fit.
Good morning. You provided your “Economic and fiscal outlook November 2015” at the tail end of November. Has anything significant happened since then that would give you a different result if you were writing the same report today?
We have not gone through things that way, so it is hard to give a firm answer to that question. We have had some GDP revisions after looking at what has happened to GDP growth in the past—the Office for National Statistics has shown a slower growth picture. We will have to take that into account, but it is not something that typically makes a big difference to our assessment of how much spare capacity there is in the economy, because we assume that it is a revision to both the potential of the economy and its actual level of activity.
Of the key things that could move by the time that we get to the next forecast in March, always the most important between the autumn statement and the budget is what we get in self-assessment income tax and some other receipts in January and February.
Some people were surprised that we did not revise up our forecast for the deficit this year by more than we did. We actually reduced it slightly, leaving aside the issue of the new treatment of housing associations. We think that there are good reasons to expect the deficit to look better in the first quarter of this calendar year than it did in the same quarter last year, relative to the previous three quarters. There are a few reasons for that. One is that past policy changes that should boost self-assessment income will show up in January and February. Another is that the change in the UK’s stamp duty regime was implemented in December and there will therefore be different impacts in the pre-December and post-December periods. Another reason is the announcement by the UK Government in June of some in-year public expenditure cuts, which have not shown up in the numbers yet. If they are delivered, we expect that to improve the numbers.
There are also a relatively small number of tax receipts that the ONS has said it will include in the official measures of receipts. It has not yet done so, but we have included them in the forecast because we are trying to forecast the numbers as the ONS will eventually define them. For all those reasons, we think that there should be a bigger improvement in the budget deficit in the fourth quarter of the fiscal year compared to the same quarter last year, relative to the previous three quarters. However, there are big uncertainties; I cite the uncertainty about how much will come in through self assessment as a key one for the January and February period. The situation is further complicated by the fact that there are policy changes that affect it, as well as what has happened with financial sector bonuses, which also have an impact.
I want to move on to the devolved taxes forecast, which you have had a couple of questions on already.
My first question concerns income tax. On page 11 of your document, you project a fairly steady increase in income tax, which I suppose is linked to growth in the economy, in wages and so on. You then make the point that the Scottish share of income tax is on a downward trend—table 2.3 on page 11 shows what is happening between 2012-13 and 2020-21. From 2013-14 to 2020-21, it goes from 2.91 per cent to 2.87 per cent, which is a fairly consistent drop. However, in the single year from 2012-13 to 2013-14, there is quite a big jump—it is seven times bigger than any of the other drops. Do you know what caused that? If that jump occurred every year, it would be extremely alarming.
The data for 2012-13 became available only in January 2015. Because the information is based on the survey of personal incomes—SPI—it comes in after a considerable lag. Fortunately, that lag will go once people are flagged as Scottish or non-Scottish taxpayers. The main reasons why you see changes in the share is that you have had a series of policy changes that are, in effect, giveaways at the relatively low income levels—particularly the increases in the personal allowance—and there have been, on the other hand, measures that increase the income tax that is paid at the top end. Because of the differences in income distribution between Scotland and the rest of the UK, policy measures that are a giveaway at the bottom and a takeaway at the top are likely to shift the Scottish share down, because fewer people pay the increases and more people benefit from the reductions. Therefore, I presume that what happened in the figures between 2012-13 and 2013-14 is because of the concentration of such measures in that year, although I cannot remember precisely which ones took effect in that year.
Will you get the official data for 2013-14 this month at some point?
We will certainly get it between now and the time when we do the next forecast. At that point, we will have another year of SPI data and another estimate of the share. It is important to bear it in mind that because it is a survey, there is going to be uncertainty about whether that information is captured. The movement in the share between any two years might be affected by the fact that the survey gives us a slightly higher measure in year 1 and a slightly lower measure in year 2. There will therefore be some additional volatility, which we hope will be removed once we no longer have to rely on the SPI to give us shares, and we are relying instead on HMRC flagging people.
10:30
That is helpful. Thank you. Moving on to LBTT, the convener has asked you already about some of the disparities between the OBR’s projections and the Scottish Government’s projections, and you have given a number of answers pointing to slightly higher house prices over time, quite a big difference in transaction forecasts over time and different treatment of the buy-to-let sector. I guess that explains what happens over the course of the forecast period.
If we look specifically at next year—2016-17—you say in table 3.3 that for the next financial year residential LBTT will be £253 million, whereas the Scottish Government says that the figure will be £295 million. I understand that by the end of the forecast period the factors that you referred to would have come into play, which would explain the disparity. However, is there an obvious explanation for why there is a difference of £42 million just for the next financial year for residential LBTT?
The difference might reflect the most recent outturn data that each of us had available at the time we were doing the forecast, and how much of that we pushed through into future years so that we could have a baseline difference. The size of the difference differs if we have the variants moving in different directions; that is one possible explanation for the difference between our figures and those of the Scottish Government.
I do not know whether the differences that we have in the amount of assumed forestalling will be having an impact by 2016-17 or whether that is mostly showing up as a 2015-16 story. I think that we have a higher number for forestalling than the Scottish Government has for residential LBTT, and I think that we also have one that the Scottish Government does not have for non-residential LBTT. That, too, might affect the year-on-year comparison. The difference might well be because we are working from different starting data in terms of what is coming in through the year, and pushing that through the rest of the forecast.
Thank you. You gave explanations about the differences over time for residential LBTT. However, in terms of non-residential LBTT, I guess the position is reversed in a way, because the Scottish Government is less optimistic than the OBR is for next year. Your projection for non-residential LBTT for next year is £243 million, but the Scottish Government’s projection is £220 million. Over each of the next five years, you are predicting more than the Scottish Government is for non-residential LBTT. Are you able to explain the difference between the figures? Are you taking different approaches?
Again, there are somewhat different methods used. I think that the Scottish Government uses our determinants to drive its forecast, but the starting point is a three-year average of the outturns rather than the latest year. Because the OBR is using a microsimulation model, we use a single base year of 2013-14, then ask how that would look if we pushed it forward into the future. That is a difference, but I am not entirely clear how much it explains the difference between the Scottish Government’s forecast and ours. Scottish Fiscal Commission colleagues might have a view on that.
Okay. Just to be clear: you are using a base year but the Scottish Government is using a three-year average.
That is right. With such forecasts we always have to choose whether we want something that is more stable or something that is more timely. There is no right answer to that, but given the techniques that are involved in a microsimulation approach, it is likely that we will lean in that direction.
Thank you. Questions on the landfill tax have been dealt with, so I will move away from the paper on devolved taxes and go back to your executive summary of your general economic outlook. Again, a number of questions on that have been dealt with already, but I have a couple of questions regarding table 1.1, which takes up most of page 12 of your summary.
Under the heading “Expenditure components of GDP”, the third component is “Business investment”. We can see, with 2014 as the starting point, that the figure goes up from 4.6 per cent to 6.1 per cent and then to 7.4 per cent, so we are clearly seeing fairly consistent and healthy growth of business investment.
When we get to 2020, however, we see that there will be a pretty sharp fall back down to 4.5 per cent, which is smaller growth than we saw in 2014. What will happen in 2020 to business investment?
I suspect that that is a consequence of the changes in the public expenditure plans. In the past year, the Government introduced a sharp increase in public sector capital investment, but there has been less of an increase in the overall public services spending. The figure may be a result of the fact that you start with a path for GDP as a whole, and then, if there are movements in the composition of the fiscal giveaway and takeaway in that year, you would assume that, because 2020 is right at the end of the forecast, that will come out in the wash in terms of the overall impact on the growth rate of the whole economy. The Bank of England will take that into account in setting interest rates, but would not take into account changes that happened in the very near term. It would change the composition. If there is more going on in the public sector at that level than you have had in the previous forecast, you are, in order to get everything to add up, making movements in the other elements. I suspect that that is the case, rather than the particular view that there is something that businesses need to be terribly worried about as we move into spring 2019.
Thank you. My last question also refers to table 1.1, slightly further down under the heading “Inflation” and the sub-heading “CPI”. Are the figures that you have set down Bank of England projections or your projections for the consumer prices index?
They are our projections for CPI. We still, in effect, assume that the Bank of England is aiming to achieve, at the end, the inflation target of 2 per cent that it has been given. We have inflation returning to target slightly more quickly than we did back in July; that is partly because of greater assumed pressure from unit labour costs pushing those projections up.
The Bank of England—if memory serves me right—has inflation returning to target somewhat more quickly than we do. However, if you look at the differences, you will see that we are talking about a tenth of a percentage point below the inflation target, so again I would not overstate the significance of any differences there. Those are our numbers. We take the market’s assumption of what is going to happen to interest but we do our own forecast for inflation.
As ever, most of the questions have been asked, but nevertheless—seeing as you are here, Mr Chote—I will go through some of the detail. I am absolutely clear on what you are saying about the Scottish share in tax revenues declining as a result of there being fewer higher-rate taxpayers and of changes being made regarding policy decisions.
Is there anything else underlying that? I am conscious that the employment and unemployment rates are different, and that there is population decline in Scotland whereas the population across the UK is expected to increase. I wonder how those other factors might come into play.
You are right that there are such differences between the employment and unemployment rates. What would matter in terms of the share would be whether the size of those differences changed very much over time. If the relationship between those things is relatively stable, even if they are different, the share will not move around a great deal, so that is not an important driver of what is going on with the share and the forecast. We are not making any implicit or explicit assumptions that the performance of the Scottish labour market relative to the performance of the labour market in the rest of the UK is materially different.
What about Scottish GDP levels relative to UK GDP levels? What are your predictions for those, or do you not do that at all?
We disaggregate our forecast for the economy by different types of spending and different types of income, but we do not disaggregate it geographically or by different industrial sectors. Other forecasters do that, but it is complicated enough to do it in the way that we have to do it.
An issue that has come up before in discussions of the Fiscal Commission’s job, our job and on what one should base the Scottish GDP is the availability and timeliness of the data needed to produce a full-blown Scottish macroeconomic forecast. It is more difficult to do that than to produce one for the UK as a whole, and the task of simultaneously producing a Scottish forecast that is consistent with the UK forecast is greater still. Therefore, we have not made any explicit or implicit judgment about the relative performances of Scottish GDP and rest-of-the-UK GDP over the forecast.
That might be of interest as we move forward.
As a side issue in your report, you talk about the reclassification of housing associations. You may be aware that we have had discussions with Eurostat and the ONS about the classification of private and public sector capital projects. Is that on your radar, or would it be more properly considered by the Scottish Government? It will have an impact.
It is not something that I am aware of. However, if you think that we should be aware of it—my colleagues may well be aware of it—I would be happy to look at it.
The housing association stuff that you mention underlines the fact that what Eurostat says is appropriate matters as much as what the domestic authorities say is appropriate. Often, such changes are a reflection of the interpretation and implementation of rules that come up at the European level. We warned people about the reclassification of housing associations back in July, because the decision to reclassify is based on where control is effectively exercised rather than on ownership and because the Government was, in effect, telling housing associations what rents to set among other things. We warned that the ONS might want to look at that and, indeed, it has. The Government has said that it wants to undertake a liberalisation that would move things back in the other direction, but we have not assumed that in our forecast. If the ONS indicates that the Government has done enough to make that a likely outcome, we will address that. However, for the time being, we are assuming that the housing associations will stay in the public sector for the duration of the forecast.
Can you tell us the order of magnitude of the impact that that will have on the budget and on public sector borrowing? We are grappling with the same issues here. Classification is at the heart of this, and discussions with the ONS continue. What are the consequences of having housing associations reclassified as a public sector project? I do not think that those are fully understood.
On the housing association side, it has pushed up both net borrowing and net debt. Net borrowing has increased by about £4 billion or so, diminishing over time, and the net debt change is about £60 billion, or 3 to 4 per cent of GDP. The business model, as it were, for housing associations is their grant income and their trying to run an operating surplus on the properties that they own. They then leverage their income by borrowing in order to build more houses.
One reason why we have seen the impact on the deficit declining over time is that the decision to restrict rent increases, which means less income for housing associations and less ability for them to leverage and to borrow more, has had less impact on the budget deficit. In addition to changing the rents that housing associations are able to charge, the Government restructured the grants to housing associations quite a lot in the autumn statement so that they are lower to begin with, higher towards the end of the forecast and skewed more to shared ownership than to standard social housing. One of the uncertainties surrounding this new element of the forecast is that the Government is, in a sense, trying to push housing associations towards a new business model that is different from the one that they probably thought that they were pursuing. It remains to be seen how willing or able they are to be pushed in that direction, and that will feed back into the size of the changes.
10:45
That is helpful. Although in a different context, we face similar discussions and issues.
I move on to the devolved taxes. There was criticism of your being overoptimistic at the beginning of the process. I am curious, because you are now less optimistic than the Scottish Government. What happened in between? I heard what you said on the assumptions made about prices—
That is exactly the pattern that we would expect to see over time. If one of us is more optimistic than the other continuously, there would probably be more for you to be concerned about.
I look forward to you both being on the same page—that will be interesting.
You said that the Scottish Government had made assumptions about prices and the number of transactions that are greater than suggested by your model. Would that be fair?
That is right. As I said, the transactions difference is larger than the prices difference. The Scottish Government is basically taking an approach—we have taken it for some time, too—of assuming that transactions will get back to a historical average. We wrestled with the matter before it became an issue when LBTT became a separate tax. Over time, we have seen changes in housing tenure and a big reduction in housing transactions relative to what we saw prior to the crisis. In those circumstances, a big challenge in producing a forecast is what the medium-term new normal will be to which you think you will be returning. Obviously, there would be a lot of uncertainty about that and both we and the Scottish Government have to make judgments about it. As I said, a difference at the moment might be that we have explicitly assumed that the normal level of transactions to which you will return will be lower than the level that we had assumed in July, because we took on board the evidence that if there were more buy-to-let properties, they would be bought and sold less frequently.
In this area, the challenge in reaching a judgment on who does the better forecasts is the length of time that would be needed to distinguish between luck and judgment when so much is dependent on the level of transactions. In the UK context, an awful lot also depends on the relative movement of high-value to low-value properties. That would be an issue in Scotland, too, but not to the same extent because a smaller proportion of the revenue comes from relatively high-value properties. Transactions are probably the larger issue, but in any event I would not put a great deal of faith in anyone’s firm forecasts for transactions.
The one thing that we have discovered is that people’s behaviour cannot be accounted for; I am sure that we will pursue that issue with the Scottish Fiscal Commission shortly. The degree of forestalling that we saw exceeded your estimates. I think that it exceeded the Scottish Government’s estimates, too. Indeed, it was only when data was available that we were able to estimate that. To what extent do you try to model for people’s behavioural responses to tax changes? I do not think that we do that to any great degree yet in Scotland.
We always try to do that because, as you said, it can be surprising how big the numbers can get. In the UK context, the classic case was the movement in the higher rate of income tax from 40 to 50 to 45 per cent. We are dealing with a part of the population that is able to manage its financial affairs and move its income from year to year, so if such announcements are pre-announced the sums of money that are moved over those periods are enormous, which makes the task of working out the underlying behaviour of income tax over those periods extremely difficult. That is still being debated now.
On LBTT, we have raised our estimate of the amount of forestalling from £20 million to £30 million. Fiscal Commission colleagues will correct me if I am wrong, but I think that the Scottish Government’s range was from £12 million to £37 million, and it initially thought that the amount would be towards the lower end of that. I think that the commission’s report on the draft budget, if I read it correctly, was pointing at a level in the low £30 millions. However, I may have misinterpreted that, so please check. We also have £10 million of forestalling effect in the non-residential forecast. I do not know whether the Scottish Government has such an estimate at all.
There is a separate issue about whether the move to LBTT and the higher rates at the top will have an impact on behaviour in the long term above and beyond what goes on with forestalling. There, we assume that the higher average tax rates towards the top will lead to fewer transactions and have an impact on prices as well. We incorporate those effects and explain the numbers that we have used, but again the uncertainty around them is significant.
Is the £30 million forestalling figure that you have arrived at the final figure, or is there likely to be further discussion?
I do not know about further discussion. More data will certainly come in over time. What we have is our best estimate. This is a much larger and more difficult example, but on the income tax rates, we go back and look at our forecasts after a few years and interpret what was going on in the underlying picture and therefore the unexplained bit that is best explained by what was going on with forestalling. That can change well after the event, so I would be surprised if this was the last word—or the last number—on the subject.
Thank you, and I thank committee members for their questions. I have a couple of further questions before we wind up the session. On resource departmental expenditure limits, you say in your executive summary, in paragraph 1.36:
“Taking account of expected underspending against the Government’s plans, we expect RDEL spending to be cut by £10.4 billion in real terms by 2019-20.”
What is the expected underspend?
It is a relatively small number. With departmental expenditure limits, the clue is in the name—they are limits that the Government sets in aggregate on spending, and historically, even when public expenditure is rising quite rapidly, departments and the Treasury tend to come in below those numbers. The departments know that the Treasury is standing there with a big stick and the Treasury is obviously keen to ensure that the limits are kept to.
Historically, we have had relatively large underspends such as £6 billion a year. Further out, the numbers are typically in the range of £3 billion to £4 billion a year. I can probably find precise numbers for you later, but they are usually in the low single-digit billions.
We are allowed up to 0.6 per cent of resource and 1.5 per cent of capital. I just wondered what the scale was, because you specifically mention that you expect it to impact on the resource limits.
In the table on page 127, we show the assumed underspending against the limits in each year of the forecast. In 2016-17, we assume underspends of £1 billion on current spending and £2 billion on capital. By 2020, the figures are £1.5 billion on current spending and £4 billion on capital.
Thanks for that. We will find out over time the accuracy of your forecasts relative to the Scottish Government’s by looking back at what was forecast and the outturn figures.
We will all be dead before we have a firm answer to that. [Laughter.]
You have morbidly reported that older people are tending to pass away in greater numbers, but you are not in that category yet, so you should not be too pessimistic.
To what extent are your forecasts for the devolved taxes based on the available outturn data?
Each time we come to do a forecast, the latest available outturn data is an important input. Typically, we look at how it compares with the forecast for the year as a whole and what that implies will need to happen over the remainder of the year for the original forecast to look sensible. It is helpful when that data is relatively timely and frequent. I think that there are monthly numbers for LBTT, whereas the data on landfill tax is less frequent and we have less to go on. It is always an important input.
It comes down to our job and the Scottish Fiscal Commission’s job in judging reasonableness. Each time we come to do a forecast, there are broad questions of methodology, but then there is the question of how we interpret what has happened from recent outturn data. If the figures are higher than we thought, do we expect that good news to persist and put it into the remainder of the forecast, or do we say, “This is noise” and assume that the figures will come back down to the level that we anticipated or maybe that there will be an undershoot next year to offset the overshoot this time? That is a key judgment about which reasonable—and unreasonable—people can always differ.
Fascinating stuff. Are there any further points that you would like to raise with the committee before we wind up the session?
No. You have covered the territory admirably.
Thank you for your contributions and your responses to our questions again today, Robert.
I suspend the meeting. We will restart at 5 past 11.
10:55 Meeting suspended.Next
Draft Budget 2016-17