Treasury Committee
Oral evidence: Budget 2018, HC 1606
Wednesday 31 October 2018
Ordered by the House of Commons to be published on 31 October 2018.
Members present: Nicky Morgan (Chair); Rushanara Ali; Colin Clark; Stephen Hammond; Stewart Hosie; Alison McGovern; John Mann.
Questions 1 - 79
Witnesses
I: Robert Chote, Chairman, Office for Budget Responsibility, Professor Sir Charles Bean, Member, Budget Responsibility Committeee and Andy King, Member, Budget Responsibility Committee
Robert Chote, Professor Sir Charles Bean, Andy King
Q1 Chair: Welcome. Thank you very much indeed for being here for the first of our post-Budget scrutiny sessions. I am going to ask you all to introduce yourselves, for the benefit of those not in the room, and then we will get underway. Robert.
Robert Chote: I am Robert Chote. I am Chair of the OBR.
Professor Sir Charles Bean: I am Charlie Bean, Budget Responsibility Committee.
Andy King: Andy King, also on the committee.
Q2 Chair: Lovely. Nice to see you back again. I am going to start by asking the question, Mr Chote, that I think I have asked you before and I think my predecessor has asked you at each fiscal event. Have you come under any pressure from Government officials to alter your forecast?
Robert Chote: No.
Q3 Chair: Thank you. In your forward to the forecast, you set out across two pages a very detailed account of the timetabling difficulties you faced from the Treasury in relation to producing this particular forecast, and you describe the forecasting process this time as “unusually challenging”. Have you had to write a forward like that to your outlook before?
Robert Chote: No, we haven’t. I think this particular fiscal event was always going to be a tricky one, in terms of timetable and process, relative to the norm. You had a combination of a more compressed timetable than usual. You had a timetable that meant that the Government had to make some key decisions, at the same time as they had some other things on their plate in the European dimension. You had a relatively large change in the underlying forecast. You had a relatively large package of measures, some of which interacted with each other in ways that were quite complicated, and the Government had the objective of delivering a target for the cyclically adjusted budget deficit to the nearest £0.1 billion in the target year, which in that environment is like trying to land a helicopter on a hat box, so it was always going to be trickier than normal.
The important distinction I would make is, first, this is not an episode in which the Government had information that they were withholding from us improperly or in order to make our life more difficult. It was an environment in which it took longer than it would have done in an ideal world to take the decisions that needed to be taken, ultimately, and to ensure that we had the material to deal with that.
Taking a slightly longer perspective, it has to be said that the fact that we are grumbling—as we did do in the forward—in one way shows you how far the process has improved since the pre-OBR years. We are holding it to a higher standard and obviously a much greater degree of transparency, so we have moved a long way from policy being made at the printworks, which has been the case in some past fiscal events.
Q4 Chair: Absolutely. Having set out those challenges, is there anything you would like to put on record as you would like not to repeat in future?
Robert Chote: As I said, some of it was probably unique to this in terms of the size of the package and the timescale. I think one thing we will obviously want to do with the Treasury is to revisit whether the sorts of timetables that we put in place are appropriate and, also, whether there is anything we can do to help.
The Treasury has its own task of trying to work out what the implications of its package of measures will be on the final forecast, which is much more difficult when it is a big package than when it is a small one. They have that technology but it may be that we need to go over that with them so that it is easier for them in future.
Q5 Chair: In the forward you set out various deadlines that got missed. Were there any particular deadlines that caused you and your colleagues real difficulty in this process?
Robert Chote: It is really about the point at which you are notified of the major measures that affect the economy forecast and the final scorecard, which allows you to get on with the process of finalising the documents. As I say, the difficulty was getting the decisions at that point.
There was a separate issue particularly around the package of Universal Credit measures. That was a package of measures that has the particular issue that the individual elements interact with each other. They interact with other policy decisions—for example, on income tax—and simultaneously you also have the Government taking a decision on the speed with which you migrate people over to the new system, so in an ideal world we would have had more time to think about the coherence between that timetable and the set of measures that were there.
Also, of course, in any welfare spending package, in a tax package the Government can think about what they want to do, get HMRC to come up with a costing, talk to us and then do what the hell it likes. In the case of welfare packages, there is obviously also DWP to be engaged with at a political and analytical level as well. Therefore, as we discovered back in 2015, there is a particular issue about whether you can get all of that looked at properly in time to do that.
As regards that set of messages, it is not really a lesson but obviously for the next forecast we will be asking fairly soon for people to get to work on checking that those numbers do look okay. There is no reason to believe, a priori, that there is a bias in one direction or the other but it needs going over carefully.
Q6 Chair: You mentioned 2015 there, and I think you talk about it in the forward as well, changes related to welfare or welfare policy announcements are a particular danger spot. It is the welfare nature of it and the fact it involves DWP that can cause particular issues in relation to timing but, ideally, you would like more time.
Robert Chote: That is true. It is also, as you know, the issue from the “Welfare trends report” that we produce that the underlying forecasting task with UC is particularly difficult anyway because you are transitioning from one system to another. You have a stable basis for producing forecasts when you are at one end and a stable basis at the other end, but, as people move from one to the other, it is very hard to know in real time what is going on.
It is also a reform that involves a lot of gross giveaways and gross takeaways of different varieties that have an apparently modest impact but, therefore, the uncertainty around that is greater than the relatively small cost versus the legacy system would imply.
Q7 Chair: Mr Bean, you have obviously got a long, distinguished career with the Bank of England behind you as well. Presumably the Bank is often asked to prepare all sorts of work and forecasts at fairly short notice. How does this process, which has happened over the last weeks and months, compare to anything you experienced at the Bank?
Professor Sir Charles Bean: It is worth saying it is a very different animal. The Bank forecasts are all essentially generated through internal discussion. There may be data we need from the Office for National Statistics quickly and things like that, but we don’t have to deal with lots of other Departments or anything like that. There is no real analogue for the forecasting in the monetary policy process.
Of course, on the financial stability side of the Bank, during the crisis there was lots of fraught late night working to do with bank rescues, when you are interacting with the Treasury and the other financial regulatory authorities and so forth, but I do not think that is a natural standard of comparison.
Q8 Chair: There is a chance that the Chancellor may have to come and do this all over again in the spring. He did talk about potentially there being another spring major fiscal event on what happens on Brexit. It will not surprise you to know we are going to ask you about Brexit in this session this afternoon. Is there anything you would like to put on record, if it were to be the case that you have to repeat this process all over again for some time in March or April 2019?
Robert Chote: As you know, when the Government have a fiscal event we produce forecasts or statements to put alongside. There is not a separate timetable for that and obviously we are assuming that we are doing one in the spring, come what may. The issue will obviously be around precisely when the timing is and whether the timing ends up being non-normal in the way that this one did, because of what you are up to at that point in the year. The other issue then, of course, is whether there are measures in it or not. Last year the Chancellor did stick with his intended view of it being a fiscal non-event. Whether that is the case this time remains to be seen.
Q9 Chair: I think you have already answered this but, just for the sake of complete clarity, do you think there are any material errors in the outlook because of the Treasury’s conduct, if I can put it like that?
Robert Chote: Not material, I would say. Clearly, there is a slight disconnect between the economic forecast and the substance of the package. When the late information on some of the decisions came in that was after we had closed the economic forecast. We have adjusted the fiscal forecast as best we can for that, and I think that should be fairly robust.
The change that the original version would have had on the growth profile: our best guess is that it would not have tipped the rounding at 0.1% in the growth forecast in any given year, so in that sense I don’t think it is material. Nonetheless, it is better that you do not have those sorts of disconnects.
Q10 Chair: In the time that you have had since you submitted the final forecast—and I know it is only Wednesday and it does feel like the Budget was further ago than it actually was—has there been any other work that you have done, particularly in relation to UC, that you have now spotted anything else, now that you and your staff have had longer to reflect on what was submitted?
Robert Chote: No, not yet. You have compressed the process as well as the Government compress the process. It has been a question of allowing the staff some period of emotional, mental and physical recovery, as distinct from setting off whole new work programmes over the last 48 hours.
Q11 Chair: Yes. We are very conscious that all of us may be busy doing other things in November as well.
Finally, before I hand over, we have touched on the spring statement, a potential new event it could be. Are you proposing to update your forecast or update Brexit assumptions once the shape of a withdrawal deal is known, and, once a deal is achieved, how long would you need in order to build in revised assumptions for your spring forecast?
Robert Chote: Our forecast timetable is dictated by when the Government want to hold a fiscal event, so the question then is: how much more information do we have in time to properly incorporate that? As you see, we produced a paper hopefully trying to set out some of the thinking that we would go through in addressing that, which I think you will be better experts on than I am: how much additional information do you expect to know about the end state trade relationship? Exactly what is the new migration scheme going to look like? What are the set of consequential expenditure decisions that would accompany whatever agreements are reached on what we would like to continue to contribute to or not?
Seeing only what I read in the newspapers, my guess is that we may not be any the wiser after a Withdrawal Agreement and associated political declaration than that. You will know better than I what level of detail you, as parliamentarians, will be expecting at that point. We will have no more than you have and no more than the general public has.
Q12 Chair: Finally, your understanding is still that the outlook you will be producing in relation to Brexit assumptions, the timing would be related to the next fiscal event. It would not be for the parliamentary vote, assuming the parliamentary vote is happening before we get to a spring fiscal event.
Robert Chote: No. Whether the Government decide the timing of the fiscal event on the basis of having moved this one before some additional news—although in the end there wasn’t that much more additional news—I guess the question then is whether, for the next one, the Chancellor decides to wait until there has been some additional news before having it, but that is way above my pay grade.
Chair: Thank you very much.
Q13 Stewart Hosie: Robert, last month the OBR published a paper on Brexit and OBR’s forecasts. In that paper you mentioned the three-day week, which knocked 3% off GDP in a matter of months. You did that in the same paragraph as discussion on a disorderly Brexit. Was there some meaning behind including these two things in that way in that report?
Robert Chote: The point it was trying to make is that talking about a disorderly Brexit would be a relatively unusual sort of event, potentially. Obviously, there are many different ways a disorderly exit could look like but it would seem likely to be a combination of a negative shock to demand, to the amount of spending and activity in the economy in that sense, and to supply capacity, the ability to produce goods and services, get them distributed together.
That is something for which it is very hard to find historical precedents that you can easily point to. I think the inclusion of that was less to say, “This is an obvious parallel and look at the size of that”, but to imagine ourselves sitting here in November 1973 and you said to me, “Can you possibly calibrate the impact of a three-day week?” The chances of managing to do that accurately I think would also be quite difficult. I don’t know whether Charlie has anything to add.
Professor Sir Charles Bean: The one thing I would add to what Robert said. He talks about the 1974 parallel being both a negative demand shock and a negative supply shock. The reason it is potentially interesting is because that was also a period when there is essentially quantity constraints operating. That potentially may happen again in the event of a disorderly no deal.
It is worth stressing there is a spectrum of no deal outcomes that are possible from what you might think of as a relatively acrimonious no deal at one end, to something that is not too far away from what is already in the Withdrawal Agreement without the last 5%, say—call it a no deal deal—which might be really quite orderly in process. Until you know what the no deal looks like, it is very difficult to even start talking about economic effects. At the extreme of an acrimonious and very disorderly no deal, it would be plausible to think that there are some quantity constraints that start biting at least for a while.
Q14 Stewart Hosie: Indeed, but that is already known in the sense that the UK Government’s cross-Whitehall analysis, post the referendum, has pointed at a 7.7 loss of GDP over the forecast period. That was based on an orderly no deal. I think what they called a mitigated WTO arrangement, so in one sense that kind of orderly no deal has been assessed in fact.
Professor Sir Charles Bean: But that is talking about the sort of longer term consequences. The remarks in the paper were really about the short-term impact in second/third quarter of next year in the event of a disorderly no deal.
Q15 Stewart Hosie: Okay, let’s move on. That unusual scenario, supply and demand shocks, is probably what led your paper also to say that it is next to impossible to calibrate with any confidence the potential impact of the sort of scenario in advance because of the lack of any relevant precedent, but the Chancellor has stated in the event of a no deal there would have to be another emergency budget. How can you approach then that next to impossible task, given the actual impact of some of the shocks may yet not be measurable?
Professor Sir Charles Bean: A key thing there is that you have to start drawing on what information you can get from participants in the economy about how they are preparing for this scenario. I would expect that in this instance they stayed particularly close to the Bank. It has its network of regional agents: 8,000 or 9,000 business contacts, the senior decision-makers’ panel of significant decision-makers in business. I am sure the bank, like us, will want to try to get a quick fix on how businesses are likely to respond.
Having said that there is bound to be a lot of uncertainty. How will households respond? The thing about these sorts of events is that you can get amplifying processes kicking in that you don’t necessarily foresee upfront—this of course happened during the financial crisis—but the way we would go about it is to try to use those additional sources of information as much as we could.
Robert Chote: Can I just add to that? How easy will it be for official statistics to pick up in real time, or shortly after the event, what is going on here? If you have the sorts of disruptions that Charlie was speaking about, it would be easier to capture some of those and identify some of those, and so the ONS’s initial view, ex post, of what has happened over that period, say of a year or two years, may end up looking very different to—
Professor Sir Charles Bean: If I could add a brief rider to that because the financial crisis is quite a good example of this. After the collapse of Lehmans, within two weeks we were getting reports back from the agent contacts that orders had fallen off a cliff and that this was a phrase that their business contacts used repeatedly. It did not start showing up in the official statistics for a couple of months or something. When the economy changes direction quickly that sort of direct information from businesses, and households if you can, is potentially very useful.
Q16 Stewart Hosie: I understand the very short-term difficulties: behavioural change. I absolutely understand that but, given the pre-referendum forecasts, of which there were umpteen, by and large one outlier had a minus 2% to minus 9% figure, depending on how we came out. They did look at budget trade, productivity, FDI, migration regulation, although not the specific consumer behavioural change. This information is there and, given the cross-Whitehall analysis with its various options and indeed the Scottish numbers, which are nigh on identical, surely to goodness when we know the nature of the deal, you will have a starting point.
Professor Sir Charles Bean: Oh, yes.
Q17 Stewart Hosie: Often between minus 2 and minus 9, or something like that—
Professor Sir Charles Bean: There are longer run effects. They will take time to kick in, but one of the reasons for putting out the document was precisely to bring together some of the studies that are out there on the potential impact on activity from various possible trading scenarios. That will certainly be a starting point that we can use to inform the longer run outlook. The question of how quickly you get to that position, though, is something that is very uncertain and then, layered on top of that, you may have the short run disruptive effects that we have been discussing.
Robert Chote: Presumably an environment where this short-term difficulty hits, that is an environment in which we will not have the information about where we are going to end up in the long-term, which would lead you to apply that sort of analysis, because we will be none the wiser.
Q18 Stewart Hosie: That I understand, which brings me to the stress test you did in light of the Bank of England stress test. Assumptions included a large fall in GDP, a large fall in value of the pound, a large structural output gap, significantly high inflation, large falls in employment, and large falls in productivity growth and so on, so you have done this. How close do you think your stress test modelling might be to, let’s say, a disorderly no deal Brexit?
Robert Chote: It was certainly not designed by the Bank in that way. I think the Bank afterwards has said it can provide you with some useful information, but that is not to say that this is necessarily a close scenario quantitatively to what might happen, although it does have the features of including bad news on supply, bad news on demand and sharp movements in the exchange rate and in asset prices.
It is a scenario—as I think the Governor has already spoken about—which encapsulates what in this event might be the particularly awkward decision for the Bank as to what you can do to stimulate demand in a situation in which both demand and supply are relatively weak. Therefore, it is less straightforward than, “Well, obviously, you would go in and stimulate demand as it was in the wake of the referendum vote in the first place”.
I think there are some qualitative features of that stress test that will be relevant to your thinking, but neither ex ante or ex post would you say, “Well, that is the order magnitude that you would necessarily be thinking of”.
Q19 Stewart Hosie: Sure. From that stress testing—notwithstanding it is not an absolute test on what might happen—what are the main conclusions we should be taking from that when we are thinking about what may need to be done, what may need to be considered, in terms of an orderly or disorderly no deal Brexit?
Robert Chote: One of the things it would tell you—and this is true of pretty much every scenario that we have put in in the back of the book—is that the fiscal consequences at large depend an awful lot on whether your short-term problem translates into a permanent loss of potential GDP and potential output relative to your expectations beforehand.
In the case of a disorderly exit, I suspect that this would depend an awful lot on whether you just gummed the whole economy up for six months but then you are back straightforwardly, or whether this is the sort of thing that has longer term lasting effects. Clearly, if it is combined at the same time with a big change in business investment behaviour, and an even greater period of uncertainty, then that could have those sorts of implications. That would be one lesson.
The other thing that we drew from that stress test—which I think is a general point and not specifically a Brexit-related one—is the outlook for the public finances in the event of a shock, it is not just of interest to look at what the shock looks like, but the fact that the public finances are reflecting the legacy of the previous crisis, are they a higher starting level of debt, a greater degree of sensitivity to interest rates and inflation surprises?
We noted that the stress test has quite a big rise in interest rates in it, which is not what everybody would expect in that sort of environment, and that has quite big implications for the fiscal position. That is what we highlighted from it in the “Fiscal Risks Report”. I think it is telling you something more generally about how the change in the public finances, from before the last crisis, would affect the channels through which any new downturn or new crisis would hit. That could be a Brexit one or it could be a financial sector one. It could be any number of potential sources.
Stewart Hosie: Thank you.
Q20 Stephen Hammond: Good afternoon and thank you for coming to give evidence this afternoon, and I apologise that I will be leaving fairly shortly because there is a Budget debate going on on the floor of the House of Commons.
Robert Chote: Timetabling is very difficult at the moment.
Q21 Stephen Hammond: Forecasting whether you may appear with any certainty or whether any effects may appear is also difficult.
In September this year, you published “Working paper, No. 13: In-year fiscal forecasting and monitoring” and you said, “On a simple comparison with the latest outturns, our in-year forecasts have overpredicted the budget deficit by an average of £6.0 billion a year”. Can you give us some flavour as to why that happened?
Robert Chote: It is a combination of reasons. To put that £6 billion in context, the budget deficit is the difference between two much larger numbers. If you add together public expenditure and receipts, it is about £1,500 billion, so I am slightly surprised it is as small as £6 billion but there we are.
The combination there of some changes—classification changes and accounting changes—have on average contributed to that but by a relatively small amount. It is also the case that over the same period the initial outturn estimate of the public finances has also tended to be revised down, so typically you discover after the event that things were performing somewhat better than you had anticipated at the time you did the previous forecast.
You then have a set of issues around the different receipts and spending issues, so I think it underlined for us how important the judgment we make and how difficult the judgment we make is on the size of bonus payments, for example, and the timing of those. I think also around corporation tax, the impact of the changes in the way in which it is accounted for, the changes in the policy related and, therefore, the broader issue of: when you get news about how cash corporation tax receipts are coming in to which accruals years and tax years did that apply?
I think one of the things that we had had difficulty with is we were attributing too much of the strength in the cash numbers to adjusting for earlier accrual years, rather than it showing up as being a strength in the near term. Another challenge I think we have had over this period is departmental under spending relative to budget, and that has tended to be pushing it in one direction as well.
I think the lesson was that you had a variety of factors, more of them on receipts than on spending. Also there are particular years—what tended to happen was that the size of the error had been declining from when we had started but then you had a particularly big one in 2016-17, with a number of factors moving in the same direction.
The lessons that we ended up drawing from all of this were, in particular, going back again and doing what we can on the way in which we model and the way in which we look at incoming corporation tax receipts, and thinking carefully about bonuses. A disproportionately large part of those errors comes from non-tax receipts, which is a relatively small part of the pot and, frankly, one that not many people take a great deal of interest in, and also one where you don’t have a great deal of evolving information to go on, for example, on what is going on with gross operating surpluses. I think that is the set of conclusions. Andy, is there anything you would add to that list?
Andy King: No.
Q22 Stephen Hammond: Therefore, your explanation is that a lot of it is around revisions to numbers as they come through later on. Should we take that as the reason why you have seen a trend of being too pessimistic because you see revisions later, which then suggest that the actuality of the data was not as bad as the first—
Robert Chote: It is part of the story but I would not overstate it. There is also the challenge that quite a lot of initial outturn data is in fact reflecting forecasts. As we have explained in the paper, it takes quite some time for an outturn number to actually be a fully outturn number, rather than incorporating the legacy of our forecast as well. It is not quite as clean a distinction as it looks, and I would not overstate, “If only the ONS would get its act together and come up with the correct numbers at the right time then it would be fine”. It is not as straightforward as that and that is not all the story.
Q23 Stephen Hammond: There are two parts. One is obviously the part you said that not all the story is the ONS. Presumably, as a result of this paper you have had discussions with the ONS about whether or not the data can be prepared more quickly or verified more quickly?
Robert Chote: It is partly more down to accounting treatment. There are some cases where you literally do not get good information until later, which is not necessarily an ONS issue, but we generally don’t know what has been going on in NHS trust spending until several months after the end of the fiscal year. You don’t know what has happened with school academy spending.
One particular challenge over recent years has been knowing what is going on with local government finances, so in the early years of our forecasts we routinely assumed that, because of the squeeze that was going on local government sources of finance, that they would start running down their reserves. Actually they did not. They continued adding to them. It then appeared that the corner had been turned and that they were starting to bring down reserves, and they have now gone back to putting money back in them again. Therefore, that is another challenge. It is not an issue for the ONS getting that together. It is just the information does come in late.
Andy King: In terms of local authorities, it is a very longstanding issue that the OBR, the ONS, the Treasury and the CLG have all been—we would all like more timely information, but it is coming from many sources and so it is an ongoing challenge.
Q24 Stephen Hammond: As a result of this paper, have you decided to make some methodology changes to your forecasting?
Robert Chote: As I said, the key thing is stepping up this review of how we do the modelling of corporation tax and how we reflect that. There is then an issue about what additional information you can get out of what is going on with the non-tax receipts, and thinking particularly carefully about the bonus assumptions that we make. Again there you do not have a great deal to go on, other than analysts’ reports and press reports of what is going on in that sort of territory, so thinking in that area. Obviously a lot of that work—and particularly on the corporation tax side—is ongoing, and has been reflected in the forecasts we have done here, and will go on being.
Q25 Stephen Hammond: One area where your forecasts have been optimistic as opposed to pessimistic is on productivity. Is that again explained mainly by the difficulties in measuring productivity growth or lack of information or methodology?
Robert Chote: This is the longstanding productivity puzzle that we discussed with you. The challenge there is: why on earth is this happening in the real world? It is not just happening in the UK. It is happening in many other countries as well, as Charlie will know far better than I do. Measurement issues can be contributing to this, but this is a bigger story of most industrial countries suddenly discovering the trend in productivity growth and, therefore, potential GDP growth is weaker than expected.
That contributes unsurprisingly to a serial direction of policy areas because, as the new information comes out, you constantly have to decide: to what degree is this showing you that we have moved on to a decisively new path? It is relatively unscientific. The longer the bad period goes on, the more you put weight on that as a guide to what is happening in the future. In particular, of course, the temptation early on was to say, “This must all be down to what is happening in the financial crisis. It is the gumming up of the reallocation of capital and once the banking system is back in better shape this is all going to resolve itself”.
Clearly, one of the conclusions that we and everybody else had to reach is that it is hard to still be relying on that story six, seven, eight, nine years after the crisis. You are right that this is the contrast between: in the year one we identified this degree of excess pessimism but over the longer term it shows up more as optimism. But if you are judging that over a five-year horizon, there is a relatively limited number of forecasts that we have done where you have the five years of outturn data to judge that on. As I said, there is an element in the measurement story but that is mostly it.
Q26 Stephen Hammond: Yes. Because Professor Chadha of the NIESR who has suggested that, while the measurement story did not account for the unexplained productivity puzzle for a while, more recently he suggested there is a significant measurement error occurring with about one-quarter of the forecast growth. Is that something you can confirm?
Professor Sir Charles Bean: I think it is plausible to think that a part of the productivity shortfall, compared to an extrapolation of pre-crisis trend, was about 20% below where we would have been had we stayed on that trend. That is big, really big, and maybe 10%—perhaps a little more—you can put down to measurement issues associated with the digital economy and the way that is measured. There is also a particular issue in the financial sector, the way value added is measured there.
I think it is implausible, because of the widespread nature of the productivity slowdown, to believe it is all measurement. It is probably true that there is a measurement element but putting a figure on it is quite difficult by its very nature.
Q27 Stephen Hammond: My final question, Mr Chote, I have an easy one. If I look at your book for November 2015 and I look at your current forecast for GDP at constant market prices, in 2015 you were forecasting 2018 growth at 2.4, 2019 growth at 2.3. You are now forecasting 1.3 and 1.6. Do you think the whole of that is explained by the decision to leave the European Union and, therefore, a fall in confidence in the economy?
Robert Chote: There is a lot more going on in the economy than that. In terms of the impact, if you look at the forecast that we produced in the immediate pre-referendum forecast—which would be March 2016—our estimate then of cumulative GDP growth, between then and now, was about 4.4%, if I remember correctly. We revised that down to 3.0. The latest data suggests 3.2, so that is pretty close. I would caution you that those numbers can move around and things can be revised but that judgment seems to have held up reasonably well.
There are a set of studies that people have done that try to estimate how much stronger the economy or how much larger the economy would have been in the absence of the vote. It is a doppelganger of synthetic control approach and the idea is basically that you look at a combination of economies that, together, would have shown very similar growth to the UK in the period running up to the vote. Then you say, “What does that combination do after the vote?” This is an area where different people are trying this. They have different methodologies. They have different groups of countries to look at, so I think it is very much work in progress for everybody who is doing this, but it points to 1.5% to 2.5% of GDP difference.
You would note that that is slightly larger than the change from 4.5 to 3.0-ish that we put in the forecast that has turned out to be okay. That could be down to the fact that the world economy has grown more rapidly over this period than was anticipated so, therefore, if you had known how well the world economy would do, back in March 2016, your forecast for both with a referendum passing and with a referendum not passing would both have been higher. As I say, this is work in progress. It is very early and people will be cheerfully debating this in the academic world for decades to come.
Chair: That is a very exciting prospect for all of us I am sure. Thank you.
Q28 Rushanara Ali: That is one question I can tick off my list. I will not repeat it.
Robert Chote: What will academics be doing in decades to come?
Q29 Rushanara Ali: No, what we have lost in terms of growth from the EU referendum, the one that you have just partly addressed.
Just on the budget speech, the Chancellor predicted that when a deal is agreed with the EU there will be a boost to our economic growth, described as a deal dividend. Why has the OBR not included a deal dividend for GDP growth in its economic forecast?
Robert Chote: We continue to base the forecast on the same set of broad brush assumption that we have made about the potential implications of Brexit, which we put in back in November 2016. We have said ever since then that we do not have any meaningful basis for knowing what the outcome of the negotiations is going to be, in terms of the key things that would matter for a medium and a long-term growth projection: what the trade arrangements would be and precisely what the migration policy is going to look like. In that sense we have kept those there.
What is certainly true in the case of the forecast, it is a medium forecast and, therefore, it is not incorporating the potential impact of a disorderly exit.
In terms of the deal dividend, that is not something we have had to address. Basically, the Chancellor is saying, “Look, policy could change and the underlying economic environment could change depending on what the outcome is”. If by that you mean that at the moment businesses and financial markets have effectively said, “Look, there is some non-negligible possibility of a really nasty outcome here, and that is affecting asset prices, the exchange rate and it is affecting business investment behaviour”, you could be in a world where simply the removal of that immediate concern about something really quite nasty—even if you are still left uncertain as to what the very long-term conclusion is going to be—could release some pent up business investment that people have just been hanging back on and it could have a response, in terms of asset prices, equity prices rising and the exchange rate.
It is not clear to me that that plausibly delivers you a huge fiscal upside, particularly if you have a bit more growth and a reallocation towards business investment. Stronger business investment does not help the public finances in the near term because of capital allowances. It actually weakens them. You end up with a stronger growth performance in the long-term, so I am not sure that mechanism delivers—
Q30 Rushanara Ali: They were not words that you put into the mouth of the Chancellor then?
Robert Chote: Sorry?
Q31 Rushanara Ali: “Deal dividend” was not a phrase that you provided then to the Chancellor?
Robert Chote: No, that wasn’t us.
Q32 Rushanara Ali: Essentially, it is not a dividend? It is inaccurate to call it a deal dividend, is it not, from what you are saying? It may just stabilise things if there is a deal but not recover the loss that we may face when—
Robert Chote: I think when he is talking about the deal dividend he is also explicitly thinking about it in two ways, one of which is, as I say, the economy could perk up with the removal of the tail. That would not necessarily deliver you much by way of a fiscal boost.
I think his second definition is, “I have taken the decisions I have taken at the moment, which leave me with £15.4 billion of headroom against my near term fiscal mandate”. There is an environment in which if there is a bad outcome, in which case you might need to spend that in dealing with the consequences, but then if you do not have a bad outcome he could say, “I can spend that on something else”. Clearly, both of those end up with more spending, so how you would think about it in that context of this deal.
What it would do I think is sharpen the existing tension between the fiscal mandate in 2021, which he is on course to achieve with room to manoeuvre, and the legislated stated fiscal objective of delivering a balanced budget by 2025-26-ish, which you are no closer to achieving and if you spent some or all of the £15.4 billion either dealing with a bad Brexit or because you pay it out because you do not need to deal with a bad Brexit, you move yourself further away from that rather than closer.
Q33 Rushanara Ali: Yes. To put it very simply, whatever the outcome of the negotiations, whether there is a deal or not, the upshot is we are not going to recover the loss in GDP growth that you were referring to. Obviously, there is lots of debate about the precise amount post-referendum in terms of what the consequence has already been following on from the referendum. Would that be true? Would that be the case?
Robert Chote: Obviously there is a big debate over the very long term, what the implications are, whether the “opportunities” will outweigh the nearer term costs and where you end up in the long-term position. Near term, the idea that you have dodged a particularly disorderly outcome but you are still left uncertain where you are going to be at the endpoint, it would seem unlikely to me that that would deliver a positive surprise that would recover 2% of GDP relative to the path that you would otherwise expect. As I say, there may be a composition issue as much as a size issue in terms of the change in business investment.
Q34 Rushanara Ali: Moving on to the transition period: in your EFO you state that for the first time you have included a transition period in your forecast, but you are saying that the inclusion of the transition period has no impact on your forecast for net trade or GDP growth. Why would that be the case? Could you say a bit more about why you came to that conclusion?
Robert Chote: When we reached the original judgment back in November 2016, one of the conclusions that we reached, which was likely to be applicable under a variety of Brexit outcomes, is that you have less trade intensity in the economy, so less export growth than you would otherwise have and less import growth than you would otherwise have.
The decision we made then, which I think is pretty much in line with the decisions that most other people on this basis have done, is that those effects wash out. There is less trade relative to the size of the economy but the difference between the two does not change in a way that is either a drag or an impetus to GDP growth.
We have incorporated the transition period because it was part of the draft Withdrawal Agreement, for the same reason that we incorporated the EU divorce bill but, as I say, it does not make much difference to your expectations of where the economy is going overall.
Q35 Rushanara Ali: Did anybody else want to add anything?
Professor Sir Charles Bean: The one thing I guess we should add is that the other element that was put in, in November 2016, was a downward adjustment to investment. The idea there was to capture these uncertainty effects that we have been talking about. That was the main short run impact on activity. My understanding is that the adjustments that were made at that point were not particularly seeking to put in the longer term impact of less trade intensity on GDP of the sort that we were discussing earlier with Mr Hosie.
Robert Chote: The paper makes the point that the investment story has been the key to the judgments that we have made so far. As time goes by, the way in which less trade would get in the way of specialisation in the areas of comparative advantage is likely to become more important as the weakness of all the suppression of business investment becomes less so than as a separate issue about longer term dynamic effects as well.
Q36 Chair: I want to ask about austerity. You describe the budget as “The largest discretionary fiscal loosening at any fiscal event since the creation of the OBR”. Speaking plainly—I suppose almost a “yes” or “no” answer—does that equate to the end of austerity?
Robert Chote: It depends on what you mean by “austerity” and—
Q37 Chair: That is not a “yes” or “no” answer, Mr Chote.
Robert Chote: No because, as I say, it is not a “yes” or “no” question. It depends on what you mean by that. Some people will look at a relatively narrow definition of public spending to come up to that conclusion, others will look more at what this is implying for living standards in terms of the implication of welfare changes and tax changes for individuals and households’ income as well. The definition of “austerity” is not one of the benchmarks that we have been asked to test again, thankfully.
In terms of the package itself, clearly this is a large discretionary loosening. Of course, ironically, that was the case when the health announcement was made back in June, so the rest of the package, the elements that have been added in here that were not there already, is much smaller. It is a giveaway of roughly £5 billion in the near term that then diminishes to a statistically negligible takeaway in the fifth year of the forecast. It has that sort of standard near term giveaway but virtue increases over time feature, but net the health spending dominates the rest of the package by a considerable margin.
Q38 Chair: Between March and October of this year obviously, you have significantly revised upwards the projections for nominal Government consumption as a share of nominal GDP. Is that almost exclusively because of the increasing spend on the NHS?
Robert Chote: Yes, principally so. Nominal Government consumption is where the sort of day to day spending on public services and administration comes in, so that is exactly where that is coming from, so you see a compositional—the fiscal loosening does not change the growth potential of the economy over the long term. It changes the profile slightly, so it is the reason why we have bumped up GDP growth as much as we have, relative to March in 2019, because of this money coming in relatively quickly. You have a change in the profile over that but it does not make a difference to the long-term growth, therefore the total amount of growth you get over five years.
We have taken a more optimistic view of that because of the judgment we made on the sustainable rate of unemployment being a little bit lower and the judgments on participation, which means that the level of potential output—so the growth over that period but the level at the end is higher. The changes in public services spending have more of a compositional than a timing effect than a long-term growth effect.
Q39 Chair: Mr Bean, how would you answer my first question? I am not going to ask you “yes” or “no”, but let’s assume that austerity is related to departmental spending—and obviously you agree with the large discretionary fiscal loosening—if you were a Government Department preparing for the Spending Review next year, would you feel that being asked to tighten your belt significantly is over? Of course, the Resolution Foundation suggests that unprotected Departments will still see cuts in every year from 2021, and that their per capita real term budgets will be 3% lower in 2023 from 2019. Perhaps if you just explore this austerity in the context of departmental spending.
Professor Sir Charles Bean: Is the chart 4.5 or 4.6? I cannot remember. There is a chart somewhere of health spending and non-health spending. I will see if I can find it. Yes, so chart 4.6 on page 142 is the big picture story. This is real resource spending, so I think of this as current spending. The yellow line tells you the story of significant increases in the health sector. As you can see, the blue line was declining for the past couple of years. It is now flattening off, but of course that being stable does not tell you what happens to the individual Departments within that.
We know that the Chancellor announced increases for defence and overseas aid, and, if the overall total for non-health Departments is flat and some are getting an improvement, simple mathematics tell you that some others must be facing further constraints but who they are, presumably, we will not find out until the Spending Review.
Chair: Yes, exactly. That is very helpful.
Q40 John Mann: I will start with my usual question. The Chancellor forgot to say what net migration would be but he suggested 800,000 new jobs and you are suggesting here that unemployment will go up, so are we right to presume that there is going to be around 800,000 new net migrants over this period?
Robert Chote: What we do state is that the growth in employment is more than accounted for by the rise in the population over this period, if you look at that in nominal terms and migration would be an important part of that.
Q41 John Mann: I think you said it drops; net migration you suggest drops by 2023-24 to 165,000, so over a five-year period that is 800,000 net new migrants in the country.
Robert Chote: Yes. We stuck with the same set of ONS population projections, which we have used before, and that is where that is taking it. If you look at the recent outturn data, it had been coming down relatively closely with that projection until relatively recently. Recent data revisions suggest that it is somewhat stronger than that total suggests but not sufficiently for us to change that as a long-term anchoring assumption.
Professor Sir Charles Bean: One other thing to add into the mix here, which is important, is we have a higher rate of labour force participation in this forecast than we had in the last one, so some of these extra jobs are being filled by people like me who are staying in the labour force longer.
Q42 John Mann: Very welcome. I am pleased that you have confirmed there are 165 and a drop to that by 2023-24, and over a five-year period it is over 800,000. It is always helpful to know that the Chancellor would actually spell out what his policy is. He has an immigration Act at some stage to come forward too.
Robert Chote: Remember the net migration is starting higher than 165,000 and moving down to that level.
John Mann: Indeed, as we have discussed previously.
Andy King: Also remember that net migration includes children, so they do not work in the UK.
Q43 John Mann: It is always interesting how that does not ever seem to get mentioned, despite all the Brexit issues and how that might change it.
On housing, I think the Government have suggested 300,000 units a year but you don’t seem to think that that is the case. You seem to have downgraded your forecast since November. Is that the case and why?
Robert Chote: You can point me in a minute to the 300,000 number. The change in the housing completions forecast is reflecting the fact that housing turnover has been weaker in the recent past than expected, and that is an important driver of our forecast for housing starts. You will not be astonished to hear that when you have a downward revision to your forecast for housing starts it feeds through to a lower forecast for completion, so it is primarily a reflection of the recent data on turnover.
Q44 John Mann: The Government have recently, yet again, totally changed their five-year housing supply targets for every local authority. In fact, for many they have reversed them entirely, and that is in the last couple of months. What cognisance have you taken of this huge change, which will obviously impact planning consents as well as land allocation immediately in looking at whether the Government can meet their target?
Robert Chote: We have looked specifically at the impacts of the budget policy decisions on the outlook for the amount of local authority housing produced and the effect of that on housing overall, rather than on Government targets.
The change that is allowing removal of the cap on local authorities’ ability to borrow, we assume would increase local authority house building by about 20,000, but that is offset because this is partly giving money to local authorities that would otherwise have gone to housing associations. Then there is also a slightly smaller adjustment that you would want to make to crowding out of private sector housing activity, so we assume the net effect is a boost of about 9,000 rather than 20,000. That is the revision that we have made on the public sector housing provision there.
Q45 John Mann: Public sector housing, obviously the Government’s figures incorporate private sector building as well. As you have raised it, the local authority borrowing for new local authority housing, you are projecting it is going to be under £1 billion a year but the figure of £1 billion comes from local authorities having to compete. Every local authority says it wants to build lots of houses so, in fact, should that figure not be a lot higher?
Andy King: I think the £1 billion figure comes from a previous measure that has been subsumed by lifting the cap entirely. In looking at the costing for this measure, that scheme was quite helpful because local authorities have submitted bids for it so there was quite a lot of information about the plans for those who were eligible and had bid. I think the value of the bids that had come in was worth around £2.5 billion, maybe a little less, and so the costing kind of assumes that that happens but slightly more slowly than would have been the case under a time limited borrowing measure.
Then some of those local authorities that were not eligible to bid will also use the borrowing space, so essentially we have assumed that all of the bids, which far exceeded the £1 billion cap, will go ahead over five years and then some more on top. In essence, you need to add up the five years if you want to compare it with the £1 billion.
Q46 John Mann: Mr Chote, in your March 2016 outlook, on page 213—I know you will not have that with you, but I do—
Robert Chote: I have a small library on a trolley behind me that I can refer to if necessary.
John Mann: —you have a line down, Public Service Pensions, update to discount rates, and you have a figure down that is zero until 2019-20, when it is £1,970, and 2021, £2,005. I cannot see that in this forecast. It has disappeared.
Andy King: It has not disappeared. That was a policy that was introduced in the 2016 budget and has been updated in this budget. It is one of the largest. Outside of the health announcement, it is the largest in terms of billions of pounds. It relates to the discount rate that the Treasury sets in order to calculate public sector employers’ employer contributions. It was reduced by 0.2 in March 2016 and I think another 0.4 for this forecast. The overall effect is in the latest version. I think it is roughly £5 billion rising to £6 billion or £7 billion a year.
The difference between the two measures is that, in 2016, Departments were not compensated for it so essentially it imposed a pressure on what they could otherwise spend their DEL money on. This time around, the Treasury is compensating Departments because essentially those employer contributions—
Q47 John Mann: Can I stop you there? That is why it is not in because, as the Treasury is compensating, it goes to zero. However, the Government decision, is it not, is to compensate only for 2019-20 and there is no decision beyond then. So in fact, the line that you have in, in November 2016, post 2020, should be in again for every consequential year and, as you pointed out, should be higher. In fact, I think it should be about £4 billion. I am asking why it has gone. That is creating a rosy picture of departmental spending, of cuts that the Departments will have to make, because Government have only guaranteed that money in the first year.
Andy King: My understanding of this is that the money has been put into the reserve in 2019-20 but, beyond that, there are no departmental allocations because there has not been a Spending Review. The total resource, departmental spending, however, reflects that higher standard.
Q48 John Mann: Two years ago, however, you put in a figure for 2020-21, so there is an inconsistency in the OBR’s approach here, a big inconsistency.
Andy King: In March 2016, this did not change resource—departmental totals—but it was reflected in the fact that it would reduce the net cost of public sector pensions, so it was a pressure. This time around, as I understand it, some of the March 2016 £2 billion has been compensated because it was an NHS cost and was compensated as part of the overall health package. Then all of the new effect from the latest change in the discount rate is compensated in the assumed path of resource DEL beyond the Spending Review.
Q49 John Mann: The Government announcement is only in the first year. Therefore, you cannot assume—
Robert Chote: That is because there are no Government departmental spending plans later on, so you do not know what you have to compensate.
Q50 John Mann: You have put in November 2016, despite that, and yet you have omitted it now. That is an inconsistency and in terms of the amount of money, which will be £4 billion a year on my estimation—some suggest more than that, but a significant sum of money, some would say up to £6 billion a year—that is very significant because that is money that is separate from the Department of Health, where Government have made a different set of decisions and guarantees. In all other Government Departments—schools, police—that is an actual cost that has to be projected, will be borne, unless there is a change in Government policy. It seems to me that you have been somewhat hoodwinked by the Government announcement here, and that should be in your report. It affects this report and the conclusions you reach.
Andy King: If you look at the top row of table A.1 on page 231, which may be a little bit far back in the book, precisely how the latest measure affects public sector pensions spending and resource DEL is laid out there. As you say, correctly, 2019-20 has been allocated. It has been put in the reserve. It has not been allocated Department by Department. Beyond that, the effect that is in those numbers—£5.5 billion rising to almost £6 billion—is the effect of allowing that higher starting point to feed through to future years.
Anything beyond that, on how that is allocated and whether Department by Department they are precisely compensated for the effect of this discount rate measure, is obviously a policy decision for the Spending Review. At that point, you could come back and ask whether the compensation had worked or whether it was going to have further knock-on effect.
Q51 John Mann: You are saying from that line, if Government do not make an additional allocation decision of £6.035 billion in 2023-24, that money will have to be found by Departments, other than Health, which Government have already guaranteed.
Robert Chote: If they were to change the RDEL path back down again, yes, and obviously Government have said that this a provisional set of spending plans for the time being, but they have been explicit at this point that the overall DEL envelope is being topped up, as you can see, not pound for pound here, but covering the vast bulk of that, over this period, and that has been reflected there. It is obviously not allocated by Department because we do not know what the total departmental allocations are going to be at that point, so we have to come back and see how that changes in the Spending Review.
Q52 John Mann: Therefore, if they do not find that £4 billion to £6 billion each year, there are going to be some very significant cuts. It is only £4 billion to £6 billion but, nevertheless, that has an impact on all your forecasts.
Robert Chote: It stands to reason; Government have given us the set of RDEL numbers against which we make some judgments about underspend. I can confidently predict that, if Government do announce that there is going to be less money, there will be less money available.
Q53 Alison McGovern: I must apologise that, unfortunately, I am supposed to be somewhere else at 3.30 pm due to the business of the House today, so I will have to ask my questions and go.
Robert Chote: I will string out the answers.
Alison McGovern: My question is to you all. I want to come back to the modelling, specifically the information that you have on the cross-Whitehall model on Brexit. Could you talk us through the information that you have? You note in your discussion paper, which Stewart mentioned earlier, that you had received some more information about that cross-Whitehall analysis. Could you talk us through a bit more about what that is and what you know about it?
Robert Chote: Yes, we have looked at it. One of the things we wanted to do when we were producing the paper was to look at not just the Whitehall study but also the other studies that people have done on the impact and to tease out the key judgments that drive the size of the responses that you get out of this. We have taken that approach, not just on the Whitehall analysis but talking to the others, as well.
You can see from the comparisons that we have made that there are a number of differences between the different studies that highlight the importance of some of those judgments. For example, how is it that you identify and try to estimate, the size of the impact of the non-tariff barriers relative to the easier task of the impact of a numerical tariff barrier? Should you be thinking about constant or increasing returns to scale in the way that you do the model? Should you be taking into account dynamic effects or purely static effects? It was teasing those elements out. We looked at the information on the Whitehall study that has come into the public domain and asked questions about that, and drew the lessons that we have drawn.
Professor Sir Charles Bean: Not surprisingly, we had the same slides that had been leaked earlier on this year, but we were able to talk to members of the cross-Whitehall team to try to understand better the figuring that lay behind some of the numbers, and so on.
Q54 Alison McGovern: Do they keep you up to date about it? Do you go and see them?
Professor Sir Charles Bean: We do not get a running commentary from them about their latest model runs or anything like that, but what we were able to understand better was the analytic framework that lay behind those slides, which were released earlier on this year, and some of the things that they were working on. Obviously, when it comes to the point that we may need to incorporate something in our own forecast, we may need to engage with them quite heavily to understand exactly where they are now, to get more detail on some of their assumptions, and so on.
Q55 Alison McGovern: Okay. Thinking of the Whitehall analysis versus other studies, would you say there are any points of consensus among those studies? Mr Chote mentioned what weight ought we to give to the different channels that might impact on economic variables. Where would you say the points of consensus are across the models?
Professor Sir Charles Bean: What the paper that we put out shows is that there is a range of estimates across the various studies, using computable general equilibrium approaches corresponding in some cases to differences in modelling assumptions but often the differences relate to different assumptions about the inputs. One thing we did discover, through doing this work, is that the main difference between different studies often hinges not on how the models work but on the inputs that go into them, and, in particular, how big the non-tariff barriers are assumed to be after we have left the EU.
Q56 Alison McGovern: Therefore, what drives the difference in the models is not some modelling assumption or some function of a calculation, or balance or weighting, but rather the political assumptions about what will produce those inputs.
Professor Sir Charles Bean: They are not necessarily political assumptions. They are judgments or assumptions about what the world will look like rather than the way they are processed within the model. It is not to do with different elasticities associated with how consumers and businesses behave and things like that, which from our point of view is helpful because, if people come to different answers because they have different models, you have to really get into the model to understand why different models produce different things and which one you believe, whereas if it is down to people making initial assumptions about, say, how big the non-tariff barriers will be—
Q57 Alison McGovern: I feel we are walking towards the Economists for Free Trade model, in which they assume that trade barriers with the EU will not rise.
Professor Sir Charles Bean: Yes.
Q58 Alison McGovern: Could you comment on that particular piece of work?
Professor Sir Charles Bean: Yes. Whereas most of the other studies assume that, as a result of leaving the EU, we will essentially face non-tariff barriers analogous to those faced by other third countries that trade with the EU now—in some cases those other studies might say the barriers do not go up the full amount, only to 85% or something like that—the Economists for Free Trade model makes an assumption, based on their reading of World Trade Organization rules, which prohibits discriminatory behaviour. The idea of the World Trade Organization rules is to ensure that there is a level playing field across countries, except where there is a free trade agreement between one country and another.
They use the argument that the World Trade Organization rules will prevent the other EU 27 introducing discriminatory behaviour against the UK but, of course, we are potentially choosing to cease to be a member of the EU and the single market, and so on, and now take on the status of a third country. Therefore, most trade experts would take the view that WTO rules basically mean that we have to be treated in exactly the same as any other third country without a free trade agreement and, if we weren’t treated that way, other countries could take a WTO action against the EU.
That is their argument for saying that we will now face 100%, or a relatively high fraction of those non-tariff barriers. The Economists for Free Trade study assumes that actually the basis on which we trade with the EU, even if we go to WTO rules, will not change very much. That is a key part of the difference between the results they get and other results.
Q59 Alison McGovern: To be clear, would you agree that that is more—I used the word—“political”. What I meant was more diplomatic. That that is not an economic question, but a sort of legal—
Professor Sir Charles Bean: Indeed. That is right. That is why I stressed that it is input; it is not an assumption about the model or the modelling, or anything like that.
Q60 Alison McGovern: It is exogenous to the model.
Professor Sir Charles Bean: It is exogenous to the model, yes.
Q61 Alison McGovern: Can we move on the trade intensity and productivity?
The OBR has said that if there is a reduction in trade intensity following Brexit, you are not currently—I think I am right in saying—forecasting an impact on productivity. Could you talk us through that?
Professor Sir Charles Bean: It is important to separate two sorts of impact, which often are conflated in these discussions and, to be honest, sometimes the OBR’s past publications have conflated the two.
There is first an effect that is associated essentially with a sub-optimal allocation of production. Trading enables people to specialise in the things that they are good at. If you put barriers in the way of trade, that means that people have to do things that they are less good at instead of doing something that they are better at. That is what you might think of as the static effect. There are still different ways of measuring that effect.
On top of that, people believe there may also be what in the document is referred to as dynamic effects, which are associated with technical change, discovering better ways of doing things, and those sorts of things.
My reading of the literature is that the evidence on the link there is not non-existent but is weaker than some of the evidence for the static effect.
Q62 Alison McGovern: Also, presumably, fairly long term.
Professor Sir Charles Bean: Yes, there is certainly a question mark about how long the effect takes to materialise; does it take place primarily through FDI investment, having foreign ventures based there, whatever. It is a more tenuous channel, I think, than those static effects, which are associated with an inefficient allocation of production.
Q63 Alison McGovern: Can I come back to some specifics, particularly on the WTO-type Brexit that we were just discussing? It is less the long-term productivity argument but the immediate short-term impacts that we might see—for example, on food prices. There is some uncertainty about what we might expect on those sorts of goods. It is unknown what would happen with tariffs—what might happen with sterling be slightly better understood—and therefore the cost of imports. Could you say a bit about where you think the consensus view is on what happens immediately to things that we need now, such as food?
Professor Sir Charles Bean: Focusing particularly on food and agriculture is quite sensible because that is one of the areas where there are relatively high tariffs in the EU. As it stands at the moment, there are no tariffs on trade between the UK and the other EU member states. If we left tomorrow, on WTO rules, agricultural products that we produce and was sold into the EU, would have the EU’s external tariffs imposed on them. There is then a question for us about what our rates are. A lot of the modelling exercises that people have done have assumed that we at least would start with imposing the same external tariffs that we have at the moment, which are the EU tariff rates—and those rates would now apply to stuff that we buy from the EU, post our exit. There is a direct increase in costs there for both us buying agricultural produce from the EU and the stuff that we are selling in.
Now, that may not necessarily be of course where you end up because, as a nation, we could decide we are not going to have those tariffs as our tariff rates and the Economists for Free Trade group argues that actually the UK will be better off unilaterally slashing its tariffs, in which case, obviously, UK consumers would not face those tariffs. The other channel that you need to factor in, in all these trade things, when you are thinking about what is the cost of buying stuff from abroad, is what happens to the exchange rate. Very often, the movement in the exchange rate may swamp any change in tariffs. That may well be the dominant factor. So there are quite a lot of moving parts to this, some of which is policy—the tariff rates that the UK ends up deciding to charge.
Chair: I expect we will be revisiting this in the course of the next few weeks and months—
Professor Sir Charles Bean: I expect we will.
Chair: —and what was it, decades or something?—but moving on now, welcome to our new member, Colin Clark, from Scotland.
Q64 Colin Clark: Predictably, being a Scottish Member, I want to bring you to the difference between the forecast of the OBR and that of its Scottish counterpart.
Why did your estimates for Scottish income tax receipts for 2016-17 differ from those of the Scottish Fiscal Commission if you both used the HMRC Survey of Personal Incomes forecast?
Andy King: I think the difference between the two forecasts that were published, prior to HMRC having published outturn for that year, were relatively small, around £150 million on tax of well over £10 billion. There are very small differences. The Scottish Fiscal Commission used the publicly available version of the Survey of Personal Incomes whereas, because our forecast is processed for us by HMRC, and HMRC is able to use all the taxpayer confidential information, it is calculated on a very slightly different tax base.
We took very slightly different views on employment growth and on the growth in incomes being taxed, but these differences were small in the scheme of things, prior to knowing the outturn. I think the more interesting question now is that HMRC has published an initial outturn for those liabilities, which was £700 million lower than our forecast, £550 million lower than the SFC forecast, and that implies that a much lower share of UK-wide income tax is coming from Scottish taxpayers than the prior year SPI was telling us.
Q65 Colin Clark: Can I pick you up on that? In the evidence, it says that there was an overestimation of the number of taxpayers in the upper parts of the income distribution. In that there is now divergence in tax rates and tax thresholds between Scotland and the rest of the United Kingdom, would it be reasonable to say that Scotland is more vulnerable to increases in the top threshold because you obviously have overestimated the number of people paying tax and, if you move the threshold up, the number of upper taxpayers will reduce.
Andy King: That is precisely the question that we will be trying to answer with our colleagues in the SFC over the next year as more information becomes available. The tax rates did not differ in 2016-17. The overestimate in the number of higher rate and additional rate taxpayers is relative to what was reported in the Survey of Personal Incomes for the year before versus what HMRC has found by flagging Scottish taxpayers, so there is a slightly different methodology here. One methodology uses postcodes as reported in 2015-16 and one is the flagging methodology.
There could be any number of administrative or other issues going on here or there could be something going on in the real world, which was that prior to this tax being fully devolved there were lots of debate about putting up the higher rate and, in order to be flagged as a Scottish taxpayer, you have to tell HMRC that you are resident in Scotland for more than half the year. If you are a relatively high income individual with a property in Scotland and one elsewhere in the UK, writing to HMRC to say, “I live more than half the year in London rather than Scotland” is not a difficult thing to do and so, if there was some anticipatory behaviour, rather than just something to do with processing the records, then that would—
Q66 Colin Clark: The divergence could lead directly to that? I mean, that happens all across the world.
Andy King: Exactly.
Q67 Colin Clark: Even in the States, people will see what the federal taxes are so there is a risk to two economies as close to each other as Scotland and the rest of the United Kingdom, if one has a higher tax rate, and it is particularly vulnerable, isn’t there, because 1% of taxpayers are paying such a disproportionate part? You do not need many to say that their address is somewhere else and you would see a marked effect.
Andy King: That is absolutely right. Even before we knew the information about 2016-17, when we were looking, and when the SFC was looking, at the effect of raising the top rate of income tax, HMRC was looking at this behaviour and whether people would change their addresses. It is a particularly significant risk for the Scottish Government because if someone changes their address, the Scottish Government loses all of that income tax. The income tax does not leave the UK, so our UK-wide forecast is less vulnerable to this type of activity.
Q68 Colin Clark: On that basis, why do you use different methodologies from the SFC’s to estimate the Scottish rates of income tax receipts? Is that just historical, or something? There cannot be any difference.
Andy King: The differences are quite small. Essentially we use the same methodology for Scottish income tax as we do for UK-wide income tax, which involves HMRC’s personal tax model, which is built from taxpayer confidential information. It is the best that we think is available. As I understand it, the SFC’s model is very slightly different.
Q69 Colin Clark: You have now signed a memorandum of understanding with the SFC, to set out practical working arrangements, to avoid these differences. Is that the point of the memorandum?
Andy King: We have principles of engagement between the OBR and the SFC, which is the stepping-stone to a MOU. Once our shared working is bedded in, we will codify this in a memorandum.
Q70 Colin Clark: What is the point of having two different organisations? What was the reasoning?
Andy King: The Scottish budget process takes place on a different timetable, and—
Q71 Colin Clark: I imagine the OBR could have done it. Aren’t you transnational? Aren’t you independent?
Robert Chote: It is for the Scottish Government to choose those sorts of arrangements.
Colin Clark: We did. I know that.
Robert Chote: One different example is that the Scottish Government are doing more of the work in house than we are. HMRC does more of the handle-turning for us, whereas the Scottish Government—
Q72 Colin Clark: Is HMRC marking its own score sheet, if you are effectively marking what HMRC is doing?
Robert Chote: Interestingly, with the forecasts that we are doing, we have a very good working relationship with the SFC, and also with the authorities in Wales, when we are preparing these forecasts, the pre-measures forecasts. There is an opportunity across all the taxes that have been devolved to see if anybody has any new approaches on the methodological side, or if there are new data that people are aware of that we would not necessarily be aware of. That works out in that way.
Our shared view is that we do not lose any sleep over the possibility that the numbers might come out somewhat different but we do need to be able to explain where those differences come from. As Andy says, in most cases, those differences are much smaller than the uncertainty around either one of those forecasts implied by the normal uncertainties of what is going to go on with wage growth, employment growth, and so on.
Q73 Colin Clark: Does the devolution of income tax and other taxes to Scotland and Wales make revenue forecasts more complex for the OBR?
Robert Chote: One of the points you come to, clearly the greater the degree, over time, that the policy rates, rules, and so on, diverge, then the more complicated it gets because of trying to work out what the behavioural implications of those differences are going to be. In part, it depends on where policy goes.
Obviously, there are issues about what information you have available. We do not produce a separate Scottish macroeconomic forecast to drive our judgments, for example, where obviously the SFC is looking at, and is able to devote more time and resources to thinking about, what particularly might be going on in the Scottish housing market or the Scottish labour market relative to us but sharing that information is useful for both of us.
Q74 Colin Clark: What I am really trying to get to is, in the same way that Europe has accused Britain of maybe becoming the Singapore off the shore of Europe, if there is a significant divergence in tax between two neighbouring countries, two countries in the same unitary market, how do you model that risk? There must be compensation. You could end up collecting more tax in the rest of the UK and less tax in Scotland. There must be an economic risk for whatever country, of divergence, plus or minus. If the rest of the UK were to slash income tax and Scotland decided to stay with higher income taxes that would be the other way around. That would cause an enormous economic effect, wouldn’t it?
Robert Chote: Yes. The thing you are interested in there is not only is there a genuine economic effect but, also, does it encourage people who would have not changed the amount of time they live in one place versus the other, to report it in that sort of way? So you would have both genuine potential economic behavioural responses and other ones as well, which you would need to think about.
Q75 Colin Clark: Finally, does the devolution of taxes to Scotland and Wales constrain the UK Government’s ability to meet their fiscal target? Or is it just net? I am thinking of capacities of labour—and we have spoken about it already—and businesses and the growth of businesses, and how dynamic business is in different areas. Does devolution create a fiscal dynamic in that if the Scottish economy slows down, or the UK economy slows down relatively, that can affect the UK Government’s ability to hit their fiscal target?
Robert Chote: The chances of it being material to the UK’s fiscal targets are relatively low, given the constraints on how much the Scottish Government can borrow and for what purposes.
Q76 Colin Clark: Equally, if they slash taxes, it would have an effect in the opposite direction.
Robert Chote: Yes, exactly.
Q77 Colin Clark: Companies could register in Scotland, say.
Robert Chote: Yes, there would be a distributional issue as well as the overall net effect across the UK.
Andy King: The much larger effects are where in the UK the tax is paid rather than the overall. Overall, of course, if people move to where the tax rates are lower, then that reduces the average tax rate, but that effect is second order.
Q78 Colin Clark: But their fiscal deficit could end up looking worse if it works one way or the other.
Andy King: Which I think is essentially what they tried to take into account when setting the rates. The SFC did a study on what it called taxable income elasticities—how much might move out of Scotland—and it took that information into account. One thing to say about this is obviously that it is October 2018—actually, it’s November, now, isn’t it?
Chair: No, no. It’s tomorrow. You are okay, Halloween to go first.
Andy King: It is October 2018, and we are just discovering whether our forecast for 2016-17 matches outturn, so that is one thing about the devolved taxes, particularly the ones that are still collected by HMRC, that is very difficult because it is just not set up to understand in anything close to real time where the tax is coming from.
Q79 Chair: Thank you. One final question, on Universal Credit. We talked earlier on about the measures, the uncertainties and the lateness of the package that reached you. There are three measures in the Treasury scorecard, two that cost money, and there is also delay, which saves money. Can you talk us through how delay saves money?
Robert Chote: Basically, if you delay the rollout, you have fewer people who are being migrated—DWP is pushing them on the new arrangement—
Chair: Postponing the financial pain, basically.
Robert Chote: —so DWP does not have to pay as much transitional protection to people who lose as a result of being moved over. That is why a delay reduces costs.
Chair: That is very helpful. Thank you very much for your evidence this afternoon. We are very grateful. Who knows, we may be doing this all over again in a matter of months. But, for now, thank you and thank you to your staff as well. They have worked very hard. Thank you.