Welcome to the Treasury Committee on Tuesday 2 December 2025. Today, we are starting our sessions discussing the Budget of last week, and our first session is with the Office for Budget Responsibility. We had expected to have the then chair of the Office for Budget Responsibility, Richard Hughes, as a witness today. As we all know, he resigned yesterday following the publication of the Hogg review into the inadvertent release of information about the Budget last Wednesday, before the Chancellor had spoken. I want to take this opportunity to put on record the Committee’s thanks for Mr Hughes’s stewardship of the OBR over the last five years. In that time, he was of course appointed during covid, and he also had to handle the challenges of the mini-Budget. We were very pleased to reappoint him to that position in May, and we wish him well in his future career. We are also clear, as a Committee, about the vital importance of the OBR in our system and strongly defend its independence, and we will obviously have a role in appointing his successor. I should say to our witnesses that we are not planning to discuss the Hogg review today, because it only came out yesterday, and we recognise that the review stated that more work needs to be done to ensure that the OBR has more secure systems in future. So we will be calling witnesses from the OBR—whoever that may be at the time—at some point in the new year to discuss what the next stages of that are. We will not be discussing that today because we recognise that you will not have any answers—[Interruption.] Mr Josephs, did you want to say something about that?

Thank you for your words there, Chair. I did just want to say, on behalf of the leadership of the OBR—and, I know, also on behalf of Richard Hughes—that we would like to apologise to this Committee, to all Members of Parliament and to the Chancellor and the Prime Minister about the early release of the document on 26 November. We do know that this was hugely disruptive to two hugely important parliamentary events: Prime Minister’s questions and the Chancellor’s Budget statement. We completely understand the importance of those events for parliamentarians and for the public. We also really understand the work and the care—across Government, from the Chancellor down—that goes into the presentation of the Budget statement, so everyone at the OBR deeply regrets that this happened. As you say, you have the report of the investigation into what happened and how we can make sure that it never happens again. We would like to thank Baroness Hogg, Dame Susan Rice, Professor Ciaran Martin and the whole team for producing that report. You can just be clear that David and I are fully committed to implementing all of the recommendations in the report to ensure that this never happens again.

Thank you very much for that, Mr Josephs. I think we would expect nothing less. We do want to get into the detail of how you secure your information in future, but we recognise that you will not have had the chance to have enacted anything in that report, which itself recognised that more work needed to be done. I would like to welcome our witnesses from the OBR today. We have just heard from Tom Josephs, who is a member of the Budget Responsibility Committee, and we also have Professor David Miles, another member of the Budget Responsibility Committee. Thank you very much for coming. I want to start with the letter that you wrote to this Committee. It was mentioned in the forecast documents last Wednesday, and we received it on Friday. How unusual was the Budget process this time around, including the information that came into the public domain and when?

I think we felt that there was a need in that letter to set the record straight on the process by which the forecasts were created at the OBR—in particular, to remove some misconceptions about them that we saw circulating in the media. We know that this Committee has a strong, legitimate interest in our processes and how they are done, so you deserve to know things, and we were planning to wait until immediately after the Budget to release the letter that came to you. The reason it came late was that, as we all now know, there was an unfortunate incident of early access by some parties to our document, so we spent most of the immediate aftermath of the Budget dealing with that and trying to set up the panel to investigate it. Therefore, the letter did not come to you until slightly later than that. I think it might have appeared to some, reading the media, that the OBR came under certain pressure and might have acceded to that pressure on several things—for example, moving the window that we use to assess market expectations of interest rates in a direction that might have been helpful to the Government. There was some speculation over that, and there was some speculation that we had somehow helpfully come up with some extra money—some funds fallen down the back of the sofa, so to speak—at the last minute. There was also a view that the OBR’s forecasts were wildly fluctuating in the process both leading up to and perhaps after the pre-measures forecast, and that that had made the Budget process more chaotic than it otherwise would have been. We were keen to set the record straight on some of those things. Some of it is very straightforward: for example, the window that we chose to use to set market expectations of interest rates, which has quite a significant impact on the assessment of where debt interest costs are going into the future, had been decided on, I believe, back in July. There was no sense in which that was shifted at the last minute to try to find a window that might be particularly helpful. It was also important to set out that in the lead-up to the final pre-measures forecast, which went to the Chancellor right at the end of October, there had not been wild fluctuations in our assessment of the outlook. They had gradually got a little bit more favourable, in the sense that the margin—the headroom, as people call it—looked as though it dipped from slightly negative to ever-so-slightly positive, but there were no wild fluctuations in those things.

We have seen that, obviously, in the letter and the table; I should say that that information is in the EFO every year anyway—or most years. What was new in the letter that you sent us, just so that people are clear?

It is a matter of record in the EFO, as you are well aware, Chair, that we show what the pre-measures estimate of the so-called headroom is—

What I am saying is that some of that information is not new—it was just put together in a letter.

Quite so. What was new was just to show the preliminary forecasts that led up to that number—

Rounds 1 and 2?

Correct. That had improved by relatively small amounts as we went through the rather long iterative process of producing the forecast in the lead-up to the 31 October number—there were no wild fluctuations there, in fact; by the standards of history they were quite small.

So really it was 1 and 2 that were the new bits. I have some quickfire questions, if I may. Did you or anybody on the Budget Responsibility Committee complain to the Treasury about any leaks of information?

I don’t think there was any formal complaint. We were obviously aware that information seemed to be getting into the press.

There was no formal request, but did you talk to the permanent secretary or anyone in the Chancellor’s office about this—I mean “you” plural, the OBR?

I think it was clear that there was a lot of information appearing in the press that perhaps would not normally be out there, and from our point of view that was not particularly helpful.

Did anyone from the OBR raise it with the Treasury?

I think it was clear that we did not find this helpful.

Can you tell us who you raised it with?

With Treasury officials. You will appreciate that we have regular meetings during the process of producing the forecast.

When did you start raising those concerns?

Certainly during the lead-up to the pre-measures forecast, when there was speculation. Some of these misconceptions, as I have described them, were appearing in the press and we made it clear that they were not helpful, and that we were not in a position to put them right.

What response did you get when you raised these concerns?

I think there was a recognition that this was not terribly helpful.

This is all very genteel. So you raised concerns generally with the Treasury?

With Treasury officials, I should say, at these meetings.

Did you suspect where the leaks were coming from?

It was very hard to say.

Are you aware of whether Treasury officials tried to do anything to put the word out to stop future leaks?

I do not have any direct evidence of that.

When did you decide to pull together the information and include information about rounds 1 and 2 in a letter to the Treasury Committee? The text of the EFO would have been finalised some time before the Budget.

From memory, Richard produced a draft, and we looked at it and commented on it quite soon before the Budget. It could have been in the final week or two.

The hard copy would have been finalised, presumably, three or four days before the Budget itself.

Yes, very close to the Budget itself.

When did you talk to the Treasury about publishing the letter? It is very clear that it is agreed with the Treasury that this letter be sent.

The Treasury agreed that we should release the letter, but it was keen to do so in a way that did not set a precedent that this would now happen routinely—that we would show many different rounds of forecasts. We would not do that in general, but it was in the context of misconceptions that were out there. Those were unusual, and we hope they will not happen again.

Who at the Treasury agreed with the OBR that this letter should be published?

My impression was that Richard was dealing directly with people in the Treasury about that. I think it certainly went to senior officials in the Treasury.

As far as the permanent secretary? Tom Josephs is nodding.

Yes, Richard talked about it with the permanent secretary.

Did any members of the Budget Responsibility Committee or the committee as a whole consider commenting on or correcting any of the information in the public domain in the run-up to the Budget? You said that you could not comment, but did you consider breaking that normal practice?

indicated dissent.

No.

For the record, the witnesses are shaking their heads. Did the OBR undertake any leak inquiry, in your own organisation, to ensure that any of the leaks about the Budget had not come from within the OBR?

We made it very clear to everybody in the OBR right from the outset, and repeated it throughout the whole process, that we do not leak information about the Budget process. I do not think there is any evidence at all that that happened.

So you are absolutely confident that it did not come from you. You have talked about the concerns about the public debate in the run-up to the Budget in general. Some of this is market-sensitive. There is obviously a lot of politics swirling around. What were your major concerns about the information going into the public domain?

There were, as I described earlier, misconceptions—one was that the OBR itself was a source of difficulty due to our forecast gyrating enormously in short periods of time. Had that happened, that would have made it a much more difficult process than it was. It was always going to be a difficult Budget, but that would have made it more difficult. We were particularly keen to make it clear that the specific decision on the window for the interest rates was not under pressure to somehow pick a particular period that was advantageous, when gilt yields might have been unusually low.

I think our concern was that there seemed to be a misconception that there had somehow been some good news. I am not sure where that came from; it did not exist. What had happened was that the forecast for headroom had gradually improved a little bit in the lead-up to 31 October, but we were not aware of giving any extraordinarily good news after that.

I am not sure how much attention people were paying to—

I do not think there was a particular concern in the OBR that there had been overt market manipulation so that people would be misled about the state of the finances. The story that the news was slightly better seemed in part to be a reflection of an earlier perception that the inflation outlook and wages were a bit stronger. It was true that that was the case, but in our forecast, most of that had already been factored in by 31 October. Whether there was a delayed response to that in the wider world is, for me, a bit difficult to judge.

My interpretation was—others might interpret it differently—that the Chancellor was saying that this was a very difficult Budget, and very difficult choices needed to be made. I do not think that that was in itself inconsistent with the final pre-measures assessment we had made. Although that showed a very small positive amount of so-called headroom, it was wafer-thin—I think I described it in an interview immediately after the Budget as a “sliver”. The fact that it was a positive number—we are very keen to be clear on this now—did not in any way suggest that the OBR assessment was that the fiscal outlook was problem-free and that there was no case for a substantial tightening in fiscal policy. I will briefly explain why. The plus £4 billion number had not taken account of something that we absolutely were going to take account of—it was well known in Government that we were—which was the reversal in some of the welfare decisions, the decision not to cancel the winter fuel payment to pensioners and the reversal on the PIP measures and other sickness benefits. That was always going to count as post-measures because they had come after the March statement, and that was going to make the fiscal position of the order of £7 billion-worth just in itself. Furthermore, the Government had been pretty clear that they agreed with us that £10 billion of headroom was wafer-thin anyway, since it was inevitable that that would, other things being equal, turn into a negative number. That was known on the 31st when they got the pre-measures forecast. If the Government had decided that they wanted to build up the headroom to a much larger number than even £10 billion, it was clear from our numbers that that would mean that they had to find a fiscal tightening of the order of £20 billion or more.

It is certainly true that there was no immediately good bit of news in that particular window.

In my own view, I do not think it was misleading for the Chancellor to say that the fiscal position was very challenging at the beginning of that week. I do not know whether a message was then put out to say it was less challenging by the end of the week, and I do not know where that message would have come from. It certainly did not reflect anything that was news from the OBR being fed into the Government.

Are you saying that the Treasury briefing that good news received from the OBR had driven the decision to reverse the proposed income tax increases was incorrect?

I honestly could not tell you what drove that decision.

The critical piece of information in your letter that everyone focused on was the final pre-measures estimate and the date on which it was released. You have helpfully clarified to the Chair already that that is published in the EFO every time.

Yes, correct.

So we were always going to find out that piece of information eventually.

True.

You said that the unusual step was to publish the prior forecasts, and you feel a need to protect them in the future. Can you explain why?

Because I do not think it is helpful for the wider world to be fed information week by week, which is happening as we iterate towards a forecast.

But what about at the publication of the EFO? You already tell us the dates.

The EFO will always tell you the date, and it will give you the pre-measures number, but to provide a kind of running commentary on all the steps that get you there, which do move around, is not particularly helpful. The reason why I think it was felt helpful to show those particular numbers on this occasion was that there seemed to us to be a misconception that the OBR was, in some sense, adding to the difficulties of this Budget through its own forecast fluctuating by enormous amounts almost from one week to the next. That did not happen.

Could you explain to us why they might fluctuate between events?

They might fluctuate as a result of, for example, an enormous increase in oil prices, a collapse in equity prices or the announcement of an enormous increase in tariffs on the UK. There are all kinds of economic shocks that can come almost in the space of a few days, and it is perfectly right that we then take notice of them and update our forecasts.

I want to know why it is not right for our Committee—after the event; not as a running commentary throughout the process—to have that information to hand.

Do you mean in real time or after the event?

No, I mean on the publication of the EFO. You already publish the dates of the different forecasts, and we could have the figures that sit alongside, with explanations about why your forecasts went up and down.

I see what you mean.

Why can’t we have that?

It could be that that is a sensible way to go. All I am saying, really, is that we were not setting a precedent in saying that that is how things should be—

Unfortunately, I think you might have set a precedent. I can see that Tom Josephs wants to come in.

Maybe I could, particularly on the fiscal forecast, explain the way the process works. It is always, essentially, a very iterative process. We actually start looking at the fiscal forecasts way back in the summer, when we try to do a review of all the models that feed into the fiscal forecast, on both the tax and the spending sides. Hundreds of different models are run by HMRC and by analysts in Government Departments for the fiscal forecast. The reason why we have a number of different pre-measures rounds is because it takes some time to get through all those forecasts and really refine them. We will be given an initial view from HMRC, for example, on the tax forecast. OBR staff will go back with a lot of questions and challenge on the forecasts we have been given. We will eventually agree with HMRC to make changes, which will be incorporated in subsequent rounds, and we go forward on an iterative process until we reach what we think is a central pre-measures round. There are a huge number of different factors, some of them very small, some of them much bigger, that will drive change at each of the pre-measured forecasts.

I want to focus on the interest rate window. You said that that did not move, and it was set early in the summer. When was that? Can you clarify what the interest rate window was?

We decided it in July—we set the window, and it did not change after that. I think I am right in saying that it was the 10 days up to 10 October for the economy assumptions about where gilt yields and market expectations on Bank of England interest rates would move. It was a couple of weeks later—I think the window in the 10 days up to 21 October—for the fiscal forecast. The reason why the fiscal numbers had a later window than the economy numbers was that the economy analysts in the OBR need longer to work through all the implications of interest rate movements. On the fiscal side—Tom will correct me if I get this wrong—it is a little bit more of a straightforward calculation, because you know what the stock of Government debt is and where that is likely to be going. It is just a question of, in a sense, applying a new set of expectations for gilt yields and the Bank of England rate to a stable stock of debt, and then getting an estimate of the purely fiscal effects.

I can see Tom wants to come in, but to explain the question, obviously this is known to the Government, and their communications can affect some of these markets, as we have seen, and as Dame Harriett Baldwin alluded to. Do you fear that at any point during the cycle the Government communications could end up affecting the end OBR figures?

They would have had to be very clever and work out what kind of information to give on which particular days to try to manipulate the yields in a particular direction. I would be surprised if people thought that that could be done successfully, even if they had a mind to try to do it.

I want to add one thing, about why we chose to take the later window for the fiscal forecast. That was because we agreed with the Treasury that we would close our pre-measures forecast earlier than had been the case in the past, to allow the Treasury more time to finalise their policy package. Because we were closing the pre-measures forecast earlier, we had some concerns that that would mean that the window we were taking for interest rates would be out of date, essentially. That is why we decided, this time, to take that later window for the fiscal forecast for interest rates. As David said, we agreed that early in the process, and we communicated it to the Treasury in September. It could have gone either way, if you like. At this event it turned out to be slightly favourable to the Treasury, but that was not necessarily going to be the case.

Professor Miles, you spoke about how the news story broke that there had been some change to the OBR forecast. It was around about the time that it was widely reported that the income tax rate proposal had been dropped. Is the real reason for the letter that, as an institution, you felt you were being used as a political tool in the discussion?

I think it was more that there were misconceptions out there that were wrong and damaging to the OBR: that we had acceded to pressure at certain points—for example, on the window; that we had helpfully found some money at the last minute; and that our forecasts were all over the shop and were making it chaotic for the Government to make any decisions. It was because these things were both wrong and damaging to the OBR and to the process.

Professor Miles, was the scene-setter that the Chancellor held in No. 11, and the wide understanding that that meant income tax was going up, a reasonable response to the final measures of £4.2 billion that you had handed over four or five days before?

The £4.2 billion was a positive number, so the central guess as to where the margin would be four-odd years down the road was, by a tiny margin, just the right side of zero. At the risk of repeating myself, that was not a number that the OBR in any way wanted to be interpreted as, “This is very, very good news—there is no hole to fill,” as people were saying.

Do you think it was sufficiently bad news to justify the way it was presented on 4 November?

I think it posed a significant fiscal challenge to a Government that wanted to increase the headroom.

But that was not the justification given on that occasion, was it?

I cannot remember the Chancellor’s words.

She did not talk about headroom, did she?

The Chancellor will speak for herself, of course, but I thought the message was more, “This is a very challenging fiscal position.” All I am saying is that plus £4 billion in the pre-measures forecast is not inconsistent with the sentiment that this is a very challenging fiscal position. We already knew that we had to take into account the reversals on winter fuel payments and PIP, so it was definitely going to go from plus £4 billion to minus £3 billion, just holding everything else constant.

I understand your characterisation of the bits that were missed out of the £4.2 billion, but surely there is a big difference between saying, “We need some extra headroom because £10 billion was too thin” and increasing taxes by £26 billion for many welfare measures?

I am not in a position to speculate on what message the Government were trying to get across and how that might have changed. I can talk you through the arithmetic. If you start out with plus £4 billion and you know it will go to minus £3 billion, and you ask how much you need in fiscal tightening to get back to the £10 billion that had been there in March, you need £13 billion of fiscal tightening. If you thought that £10 billion was a very small number given the uncertainty, which the OBR judged was a reasonable assessment, and you thought you were trying to get £20 billion rather than £10 billion, you would need £23 billion of fiscal tightening instead of £13 billion. That is quite a challenge. Therefore, the message that it was difficult was not inconsistent with the plus £4 billion number. I am saying no more than that.

To go back to the letter, when did the OBR decide that it was going to signify in the EFO that it was going to write a letter?

Maybe Tom can help me, but my recollection was that it was quite close to the Budget.

What event precipitated that decision? You had seen weeks and months of characterisation of the OBR’s position in the media, presumably from leaks, which I think you have alluded to. What had not happened before that finally made that happen?

We were reading stories in the press—everybody could see them—that the OBR had come up with some favours for the Government; that it was under pressure; that it acceded to them; that it moved the windows. There was also a different range of stories: “The OBR’s all over the place”; “The OBR is the source of the difficulties in finalising the Budget”. All that kept carrying on. At the risk of repeating myself, the letter was to try to remove misconceptions that the OBR was the patsy that was doing what the Government wanted or that the OBR, through its own fickle behaviour and changing from one day to the next depending on whether it was sunny or cloudy, was making it virtually impossible for the Government.

But that was not the case?

No, it was not the case. Indeed, the letter showed that, in the run-up to pre-measures, that was not the case.

Due to the very unfortunate pre-Budget leak, which had huge implications, Richard Hughes is not here today, but the truth is that that letter was about something completely separate from that event. It was about, essentially, the level of trust between Treasury officials and the OBR in the preparation of a Budget for our country. Where do you think we are now in terms of that relationship? It is pretty important for the public to understand where that relationship is. It is pretty difficult to see that you have a good relationship that is set fair for the future.

I would not say that we were at war with the Treasury. We have a very close relationship with the Treasury. In fact, we rely on not just the Treasury but other Departments in Government for analysis of many sorts of measures—the DWP and HMRC, as well as the Treasury. There are lots of very good, from my point of view, economic analysts in the Treasury, and we rely heavily upon them. I hope we can run a process in the future—we will do everything we can—that is somewhat smoother than the process that we have just been through.

It is the nature of our role, as essentially a watchdog, to have what you might call creative tension with the Treasury.

Not destructive tension.

Absolutely. We are not doing our job if we are not having disagreements with the Treasury on analysis, and taking different judgments from the Treasury, which in many cases they will disagree with. That is always part of the process. I think we have, over time, built up a way of doing that while also having the positive relationship that you need to have in order to produce a Budget together. We can combine being tough—

It feels rather different this time though, Mr Josephs.

I think we combine being tough in the messages we give, and not pulling our punches, with being able to work with the Treasury constructively. We will be very much looking to do that in the future.

Finally on the letter, how did the Treasury react when they heard that you were sending it? You talked to them about it; it is clear in the letter that you had discussed it with them.

We discussed it with them, and as it says in the letter, they did agree that it should be sent.

Were they cock-a-hoop about that?

You will have to ask them about that. They agreed to send it.

Did you have to do any sort of deal that you would not go out publicly during the process, or would you have even considered going out publicly during the process?

Absolutely not. No deals.

Thank you. That was just for clarity.

Professor Miles, you have explained that the leaks and endless speculation about this Budget, which seem to have been a lot greater than for any Budget that I remember, were not helpful. You have focused in particular on confidence in the process and confidence in the OBR, with different views being cast on the OBR, but there is a real world out there of businesses and consumers who also have to deal with this endless speculation. Quickly, because we need to move on from leaks, is this endless speculation, leaks and all the rest of it damaging to the economy and damaging to businesses and consumers—as an economist?

I think uncertainty generally is not helpful for most economic decisions. People will postpone decisions. They may decide to hold off on investments. Removing uncertainty on the whole is a good thing. Of course it was inevitable, given the date of the Budget, that uncertainty would last for some months in the lead-up to it. That was almost inevitable, but it may well have been exacerbated by leaks, which some days seemed to be suggesting one thing and the next day something different. I do not think that can have helped.

So it damaged growth, all other things being equal, even in a small sense.

Yes. I think it is a bit hard to quantify because, even in the absence of any story speculating about this or that measure, I think it was pretty clear that the Budget was going to be a difficult one and that because of what had happened in the economy, and because the headroom had been very small before, it was quite likely that there would need to be tax increases. It was not as if there was an absence of bad news, in some sense, coming down the road, but I think it was exacerbated in its effect on uncertainty by stories about, “Well, it could be that measure. It could be this measure.” I do not think that helped.

In the previous budgetary process in the spring, we had the late adjustment to welfare policy scoring. It seems to have prompted a change, this time round, with the movement towards one assessment of the performance against the fiscal rules, with two forecasts. Can you explain to me what the impact of that decision will be?

This is the Chancellor’s announcement that, at the spring forecast, there will not be an official assessment against the fiscal rules. That is designed to support the Chancellor’s aim of just having a single fiscal event. We will change some of the presentation in our report to reflect that. We will continue to produce transparently a full economic and fiscal forecast over five years, which will include forecasts for all the fiscal aggregates. That will include those aggregates that the Government currently target in their fiscal rules, but we will not produce a formal assessment—if you like, a sort of “yes or no”—as to whether the Government are achieving their fiscal rules. We will therefore make some changes in the presentation of the current chapter in the report, which is basically focused on that question.

Will you come to that figure internally but not publish it?

We will publish, as I say, all the information on the public finances, including forecasts for the current balance—

My question is, will you determine the figure but just not put it in the EFO this time?

The number for the current balance will be in the EFO, and the number for public sector net financial liabilities. The two measures will both be in the book. There will not be a formal assessment of whether or not the Government are meeting their targets, but the numbers will be there.

Ordinarily, as you know, the Treasury has the OBR and economists in before we have the Chancellor in. Will the economists be able to tell us the answer?

Yes.

So in reality you are not making a formal assessment, but we are still going to have that assessment? The speculation will remain the same and the reaction to that figure will probably remain the same.

The numbers will be there. We will try to provide some more context in the report, because we will have a bit more space and time to explain some of the uncertainties around those numbers and some of the risks ahead of the autumn Budget, which is when the formal assessment of the rules will take place. The idea here, which is actually a model that is followed in many countries, is that the spring forecast would be more of a preliminary view to set the grounds for the Government to think about their policy settings ahead of the autumn Budget, when they make their final policy decisions and the formal assessment against the rules is made. We would aim to use the spring forecast document to provide more of the context, risks and uncertainties ahead of the autumn.

I understand the sentiment, and I know that the IMF made recommendations to this effect. It wanted to de-emphasise the concept of headroom in public debate, but the reality is that if they are seen to be off-track, and there is no repair made in the spring statement, the markets are going to react accordingly, aren’t they?

As I say, the numbers will be there. The Government have said that they will not respond unless there are exceptional circumstances—I think that is the language they have used. This is not an unusual way to run a fiscal calendar and a fiscal process. It is very similar to the approach used in many countries around the world, where you have a single main fiscal event where policies are changed, and a preliminary forecast ahead of that.

Richard Hughes, former OBR chair, told the Treasury Committee in July that moving to a single annual forecast would make us “one of the least fiscally transparent countries in Europe”. Do you think the fact that there are still two forecasts, but only one formal assessment, means that that statement does not hold true now?

Richard was talking there about a proposal to have just one forecast a year, which would be a big reduction in transparency. That would be very different from nearly every other country around the world, which has two full forecasts every year. That is not the proposal that the Chancellor has made.

Just one final thing on the IMF report: the IMF said in July 2025 that one of the strategies you could take is to have “Proactive communication to manage market expectations. Continuing to telegraph and explain policy changes”. Do you think that the Government listened to that particular element this time around, and if so, was that the reason behind a lot of the information getting to the newspapers early?

You would have to ask the Government about their communication strategy. That is not for me.

Nifty swerve, Mr Josephs.

I do not really see why that would be the case. As I say, the Government’s communication strategy is not a matter for us. We were just focused on producing the report that we are required to produce under the legislation, and we will obviously abide by the change that the Government are making.

Thank you both for appearing before us this morning. Mr Josephs, I wanted to go back to the point that you both touched on very briefly about the fluctuations between rounds of forecasts. From my reading, the change between round 1 and round 3—so, across all the pre-measures rounds—is about £6.7 billion. Am I right in calculating that amount between the first round and the third round?

That is right.

Do you have a sense of whether that amount is large or small compared with previous fluctuations between round 1 and round 3?

I would say that these are generally rather small changes compared with some fluctuations that you see between forecast rounds, partly because we have actually been trying to introduce more stability into the forecast through a couple of routes. One route is that on the economy side we do our supply stock takes in the summer and take the big judgment on productivity, which we took this time early. So, we have taken that judgment right at the start of the process. On the fiscal side, we have also initiated—or further developed, I should say—a process of engaging with HMRC and the other forecasters in the summer to agree any big modelling changes at that point. We try to get that into the first round, so that subsequent rounds are more stable, because we very much understand that, for the Government, trying to develop policy in an environment where the numbers are moving makes it difficult. Sometimes it is unavoidable, because there will be new data that come out on either the economy side or the fiscal side, which means that the numbers will change substantially and there is nothing you can do about that.

So in this case the £6.7 billion fluctuation from round 1 to round 3 is actually smaller than it might have been in the past, when your forecasts were more unstable. I am curious to know, and would be very grateful if you could set out, perhaps in a letter, the volatility from previous rounds that we have not been able to see, following on from Mr Dean’s comments about having this historical data, if that is—

We can discuss that.

Can I also clarify a figure from the EFO itself? You write about a “£21 billion average absolute revision” to borrowing “in the…pre-measures forecast”. Is that £21 billion the difference between different six-month forecasts? Am I reading that correctly?

That is right, yes—from one forecast to the next.

Right. So between one six-month forecast to the next, the “average absolute revision”, pre-measures, is £21 billion. That is quite a large part of historical and current—

Yes, and that reflects the uncertainty around any fiscal forecast. Obviously, over the last 15 years that the OBR has been in existence, there have been some big shocks to the economy that have changed the fiscal outlook, and there have been some big changes to policy that have changed the fiscal outlook, so it reflects that.

To go back to your earlier point about trying to address the misconceptions out there that forecasts vary wildly between rounds, is it correct to say that between different rounds, week to week, there might not be that huge variance, but between six-month forecasts there is certainly a large variance, in terms of the context of the fiscal headroom?

Yes.

I will turn now to Professor Miles on the macro picture on productivity. We have spoken to you before, when you have made previous appearances before this Committee, about your optimism on growth relative to other forecasters. Of course, the IMF, the Bank of England and others downgraded the UK’s productivity in previous years—in the last two years—and there has been a lot of discussion about your decision to downgrade it now as opposed to previously. Could you reflect a bit on that decision now? In hindsight—it is very easy to say things in hindsight—do you still back your judgment that it was right to downgrade now and not earlier?

I do—but I would say that, wouldn’t I? We are trying to form an assessment of the underlying trend rate of growth of productivity. Underlying trend rate of growth of productivity is not something where you can just look at the measured productivity, as the ONS produces it, and say, “It’s doing that, and therefore the trend is this.” The reason is that, setting to one side problems that have been particularly acute in recent years with measuring labour input, there are periods when there are major shocks to an economy. Covid and the energy price shocks after the Russian invasion of Ukraine are two examples—covid in particular. They generate huge fluctuations in measured productivity—sometimes huge reductions, and then a bounce back—that tell you nothing about the trend rate of growth of productivity. To try to reach a judgment a year or two ago—when we were still quite close to the after-effects first of covid and then of the energy price increase—would have been very difficult, because you would be trying, through a fog of noise in the actual measured productivity, to infer something about the trend path that it might follow in the future. It seemed to us that you should leave enough time to have gone past the aftermath of this temporary huge shock, particularly of covid but also of the energy price increase. There were enormous increases in energy prices in 2022, but those increases in energy prices had, to a very large extent, been reversed and in some sense had gone away by early 2023. Mid-2023 was three years on from covid’s most dramatic impacts, which were in the middle of 2020. The view we took was that, in order not to be influenced in the wrong direction by the aftermath of these big shocks, you have to wait for some of the dust from that to settle. By the middle of this year, we had had about two years of what I would think of as data on productivity that was unlikely to be seriously affected by those earlier shocks, from which you could then make a judgment about whether there was going to be a bounce back to an upward-sloping trajectory or not. It turned out that, although different measures show different things, sadly, none of the measures of productivity growth over the last two years or so showed much higher productivity growth than they had in earlier periods after the financial crisis. That, among some other things that I can describe in a moment, led to the view that we had been too optimistic in thinking that we would get a bounce back from those big shocks. In the lead-up to those shocks—covid in particular—in the period between about 2015 and 2019, productivity growth did seem to be rising a bit. It looked like it might be on a trajectory that would get back to our immediate pre-covid estimate of trend productivity growth, but that did not seem to be showing up in the two years when the dust had settled after covid and the energy price shock. That is a long-winded explanation of why I think we would have been jumping the gun and a bit too trigger-happy to reach this judgment in 2023 or 2024.

I appreciate that there is something to be said for letting the dust settle. How is this decision made institutionally? For example, I think you have around a dozen economists working in the OBR. Was there a discussion with them? Was there a discussion between the three members of the committee?

Yes, absolutely. To cut a long story short, we had three or four very good OBR economists working on pretty much nothing but this. They were looking at a range of different evidence from long periods back in history through to the most recent period, thinking about the effects of AI and how that might boost productivity in the future, and trying to factor in the more recent negative impacts of tariffs on the UK and the general de-globalisation of trade across the world. We have an advisory panel made up of academic economists, City economists, leading think-tanks, the Institute for Fiscal Studies, the National Institute of Economic and Social Research, research institutes and the Resolution Foundation, all of which have done good work on productivity. We also had quite detailed discussions with Treasury economists, who have themselves done good work on productivity. I had one-to-one conversations with two real experts on UK productivity, one of whom is John Van Reenen, who was sitting in the Treasury. The other is one of my colleagues from Imperial College, Jonathan Haskel. That proceeded from the spring right through to the summer and we reached our preliminary judgment in August, which subsequently did not change what we were going to do.

You had many conversations with three or four internal economists as well as all the external economists you mentioned. Having spent a lot of time among academic economists, I know they are prone to disagree about many things. Ultimately, and institutionally, is it down to you to make that decision? Is it you and you alone, as opposed to any other member of the OBR committee?

I take the lead on economic analysis, so it was perhaps likely that my colleagues would look to me to reach my own judgment first, then to assess whether they agreed with it or thought it was still too optimistic or had become too pessimistic. It was a decision of the OBR committee, but given that it was about an economic issue and my role is to be the lead on economic issues, I suppose I got a bit more weight than the others in the decision. Ultimately, it was a collective decision of the three, now two, of us.

Does this tend to happen on the committee—for example, Mr Josephs, being the fiscal go-to, gets the say over fiscal issues and you get the say on macro issues, and that you each keep to your own?

I would not say that you get to say, “This is my judgment; I am not interested in what you other two think.” Tom does not do that on the fiscal side. Tom knows much more about the fiscal issues and complexities of the UK accounting for Government spending here, and what is on and off balance sheet, so I listen to him with interest and I learn a lot. Maybe he learns a bit about economics from me—although I am not sure that is true.

To add to what David went through there, in terms of discussions with staff on the productivity judgment externals, I was pretty much fully involved in all that, as was Richard Hughes. As David said, although he leads on macro, when it comes to making decisions, we basically do that by consensus.

The three of you make a consensus decision. Thank you; that is helpful. I have a small point on productivity. Professor Miles, you mentioned the view that you had been too optimistic. An animal has been invoked by journalists to describe that feature, the productivity porcupine, which has spikes above but not below. That means that productivity has only been revised downwards, and that you have been too optimistic, not just in this one instance but over time. How do you respond to that criticism?

I think what has happened in the period since the financial crisis is that there has been a collection of very big shocks, which have all turned out to be negative. In the financial crisis itself, we all remember that the financial system came close to complete meltdown. The banking system simply stopped working. That was a huge negative effect on output. It did not have such a big effect on employment. There was a huge slump in productivity and quite a lasting recession after that. I do not think that has an awful lot to do with trend productivity judgments into the future, but it would certainly mean that any forecast made just before the financial crisis looks ludicrously optimistic. The same goes for covid and, to some extent, energy.

Professor Miles, we spoke about the difficulty for the Government of numbers moving rapidly. There is a major concern for me: I was looking at your dynamic policy scoring document for 2023. If I understand it correctly, the OBR does not score any fiscal multipliers from year five onwards. They assume that Government policy has no impact. However, I was looking at your productivity scoring analysis, and in your own paper you say on the literature review that on a 10% increase in UK R&D some authors find a long-term effect on productivity of 1.9% increase, which is enormous, given the Budget had a £16 billion downgrade from a 0.3% reduction in productivity. How can that be a responsible way to make economic policy, if we are not allowing the Government to take into account pro-growth measures?

I think we do. For example, back at the time of the March fiscal event, where the Government had some clear new policies for us to consider on planning and house building, we did factor those in. We got a projection of significantly stronger house building, particularly in parts of the country where there is a very clear housing shortage and houses are very expensive. We factored that in, and it did affect our judgment about productive potential in the economy at the horizon. We also said that it would become larger the further ahead you look beyond the five years over which the Government’s targets are assessed. I do not think it is quite right to say that we do not pay any attention to measures that the Government think are pro-growth. Sometimes they fall beneath our threshold of, “Is the likely effect of these things on our central forecast large enough five years down the road that it could confidently be considered significant?” That threshold is 0.1% of GDP five years down the road. That is quite a small amount; it is about £3.5 billion in a £3.5 trillion economy. The threshold does not exclude things that could have a moderate, but significant effect. Often, we may form a judgment that something is unambiguously good news—the India trade deal falls into this category—but, because it has quite a long time horizon before it builds up to its full effect, it has not reached the threshold. We do not “score” it, but we can make it clear that we think it is something that has a meaningful effect, and very likely above the threshold, just not within the five-year horizon.

Am I interpreting this document wrongly then, given what you are saying?

Can you just refer to the document you are talking about?

It is the “Dynamic scoring of policy measures in OBR forecasts” from 9 November 2023. It assumes a fiscal multiplier of zero for public investment in year 5.

The multiplier is a slightly different judgment. Apologies if we have been talking slightly at cross purposes. The multiplier judgment is in a sense an old-style Keynesian thing. It is that the Government spend more on whatever. It does not matter what—it does not have to be pro-growth measures. It can be that they cut taxes and give people more money. That is a boost to demand in the short run. Our judgment is that that boost to demand nearly always has a positive effect on output in the short run. It may have no impact on what we call the underlying potential productiveness of the economy. What it does is boost demand, but it does not really have much effect on how much could be produced in the UK. It may well take demand above what you might call equilibrium output and be marginally inflationary. Our assumption is that those effects can play out over two or three years, but ultimately the Bank of England’s role is to control inflation, which generally means bringing demand in the economy back in line with productive potential. For something that just boosts demand in the immediate short term, we take the view that the Bank of England will adjust to that, to bring the economy back to what you might call equilibrium, so that demand is in line with the supply potential of the economy. That would mean that the Keynesian demand effects—not trivial in the short run—tend to fade away if you look down the road.

Even though your own report says that it has long-term effects.

Government policies can have long-term effects if they change in a permanent way the productive potential of the economy. I think the house building stuff is in that category, and I think the India trade deal is in that category.

In your analysis, does your productivity forecast essentially still come down to a top-down judgment based on past trends, as we have discussed in previous sessions?

It certainly looks at past trends over quite long periods of history, to see what happens after a big negative or positive shock in the UK—we draw upon that. It is also not just backward-looking, because we have made an assessment of the potential impact of huge improvements in artificial intelligence. On the negative side, there is a much less favourable trade environment than had looked likely at the beginning of this year, as a result of tariffs levied by the US. It is both what you can learn from history and how you can take account of the things that look different from history, both positively and negatively.

Is it possible to build a forward-looking, bottom-up forecast of productivity? One of my concerns is that, obviously, these numbers have been so wrong over the last 15 years, and that has created a large negative impact on Government policymaking.

You are right. Certainly, when you look further down the road, it is a judgment that has significant impacts. It is also a judgment that you simply have to make, because you cannot be agnostic about it and say, “We have no view on productivity”. You at least need a number for the most plausible outcome, about which there are huge risks. One of the reasons why we did some scenarios in the report we put out is to answer the question of what would happen if productivity stayed much lower than our forecast for the long run, at about the average since the financial crisis, and the answer is that it would be bad for GDP and would make the fiscal situation much worse. There is also the question of what would happen if we have downgraded our productivity forecast at exactly the wrong time and we are about to see a period over the next 10 or 20 years in which productivity goes back to what it was in the 50 years up to the financial crisis, in which case things will be dramatically better than we think.

You announced a new approach to scoring supply-side impact policies. In plain language, is it now more difficult for the Government to get pro-growth impact policies included in the forecasts, or is it easier to avoid anti-growth impacts?

We set the threshold at quite a low level, 0.1% of GDP, so does it move the dial by about £3.5 billion, either plus or minus, five years down the road? That is quite a low threshold, so it does not rule out our taking a view on all kinds of potential measures.

It is symmetric, so it would apply both to policies that are positive for growth and to policies that are negative for growth. The reason we have done this is because we had been looking at lots of policies that were essentially very small and had very marginal impacts, in both directions, on growth at the end of the forecast. We found ourselves making very small and uncertain adjustments to what is already a very uncertain view of productivity in five years’ time. We therefore thought it made sense to be clearer and more transparent about what we mean by “significant”, which is why we introduced the threshold.

The panel has already explained how the pre-measures forecast of a £4 billion surplus is not necessarily positive news for the country, but it could have been much worse if not for the growth in labour income. Could you explain what that was driven by? How much of a surprise was it when you got to that stage?

I will quickly talk about the economics and then how it feeds into to the fiscal projections. It just looks as though wage settlements this year have turned out higher than we thought. It looks as though wage settlements are still running not far off 5%, even though the labour market has been weakening a bit, vacancies have come down and, indeed, unemployment has very recently edged up a bit. It was a surprise that wage settlements seem to be hanging in at that level. Looking at company expectations, next year’s settlements will also be a bit higher than we thought back in March. On the inflation front, it is a little bit the same story. The inflation numbers through the course of this year have been coming in a bit stronger, so we have increased our assessment of inflation for next year by about 0.5%. We have increased our forecast for hourly earnings by a bit more than that. Both of those things, particularly in an environment in which we have ongoing threshold freezes on income tax, bring in a lot of money.

It is particularly the case at the moment that the receipts forecast is particularly sensitive to changes in inflation and nominal earnings because of the personal tax threshold freezes, which just means that for a given increase in inflation and earnings you get a bigger boost to the receipts forecast than is the case if those thresholds were rising in line with CPI. There were also some other changes within the receipts forecast around the composition of growth, which is skewed more to consumption rather than to investment, which is also fiscally favourable, because essentially consumption is taxed at a much higher effective tax rate than investment. In fact, investment has a negative effective tax rate because of capital allowances. There were a few other factors within the receipts forecast that pushed it up alongside the change to earnings inflation.

My follow-up question is that, while clearly this was positive news from the Chancellor’s perspective in trying to meet her fiscal rules, some have reported this as positive news for the economy overall—there was good news in there. To what extent are these factors good news? I am particularly thinking about the fiscal drag element. That is going to reduce people’s disposable income. It might have an impact on consumption. Inflation, obviously, we are actually trying to drive downwards at the moment, not least because of the impact on interest rates and other things. Though this might have helped the Chancellor to meet her fiscal rules on this occasion, is this actually good news for the economy overall?

I think if it was just inflation being higher and nominal wages higher but not in real terms—they would grow a bit more simply to keep up with inflation—that in itself is neither good news nor bad news. It is kind of neutral, in a way. Of course, what it does, with the frozen thresholds, is it effectively increases the tax rate on incomes because people will get pushed into higher bands. If that were all that had happened, it is really a way of just extracting more tax and not increasing the wealth of the country.

One thing to add is that we are just talking about the receipts side. Higher inflation and earnings also mean higher welfare costs through benefit uprating. It also means more pressure on departmental spending budgets, which are fixed in nominal terms by the Treasury at spending reviews. Therefore, higher inflation means less real spending power. That does not have a fiscal effect, unless the Government adjust those nominal totals, but it does have an effect on real spending power in the public sector.

We could see the real effect of this, perhaps, in future fiscal events.

My constituents will welcome the action on the cost of living in the Budget, from the rail freezes to the cuts to energy bills, but could you set out the trajectory for real household disposable incomes over the Parliament? How does it change compared with your forecast in March, and what is the main reason for that change?

Unfortunately, it is a bit less optimistic. Real household disposable income per person seems to me the best measure because it adjusts for population, it is in real terms, and it adjusts for taxes being taken off people. If you wanted a single number, real household disposable income per person seems a good one to me. It is a bit stronger this year than we had thought, largely on the back of wage settlements being a bit stronger, but what happens further down the road, really, is that it is very difficult for the judgment that we have made on productivity, to reduce it by about 0.3% every year in terms of the growth rate, not to show up in lower growth of real household disposable income. Looking beyond this year, the growth in real household disposable income is around about 0.3% of 1%. In the longer-term history of the UK—certainly over the last 100 years—that is a very low number, but I think it is almost inevitable when there is, first of all, a significant fiscal tightening and secondly, we made a judgment—a difficult judgment—that the rate of growth of real productive potential in the economy is a bit lower.

Is RHDI the most accurate measure of living standards? I ask because it includes imputed rents and it is gross of housing costs. Would a measure net of housing costs not be more accurate?

The imputed rents thing can seem a bit weird.

We all know, but for anyone watching, do you want to explain what imputed rents are?

Imputed rents is the valuation placed on the services that people get from owning their own home and, in a sense, not having to pay rent. In one sense, that is a real source of income. If people just decided that you would pay rent to your neighbour and they would pay rent to you, that would show up as rent, which would be income, which would show up in GDP. Just because people own their houses does not mean that the services they get are any different. The idea, which has some merit to it economically, is that you count the so-called imputed income that people get from the rental services of a house, and the only reason why they do not pay rent on it is that they actually own it themselves.

Including housing costs? Why would those be included? Is it not a measure that excludes housing costs?

I am not quite sure that there is an obvious offset there. The real income from imputed income is a real thing. For example, if interest rates went up and people had to pay more on their mortgages, in some sense that would be a transfer. Think about a bank or building society that is getting deposits from UK households; if interest rates go up, that is a kind of transfer from those who have a mortgage to those who have savings in the bank. That is distributional, but it does not affect the average level of real income in the country, so it is more of a distributional thing. But your question is more subtle than that. I need to think about it.

Maybe we could follow up on that point.

Okay—maybe we could.

May I quickly turn to migration data? The latest data is much lower than you forecast. What are the fiscal implications of that?

I do not know quite the composition of the people who have come to the country more recently; it depends very much on who is coming. Are they coming predominantly to work? Will they earn average or maybe higher than average incomes in the UK and become taxpayers very quickly? In the short run anyway, that is positive; even if they have dependants, their dependants will not be eligible for welfare payments for some time. So the short-run impact of the great majority of those people coming into the country is positive; if those numbers go in the other direction and come down, and far fewer of those people are coming in, in the short run it is fiscally less favourable. I have not looked in detail at the very latest numbers, which only came out a few days ago, I think. It all depends very much on the composition. You could have a reduction in the number, but an increase in the proportion of people who will be directly working and far fewer dependants—and there may be far fewer students, who will not be generating any taxable income. It depends very much on two things: not just the absolute number of net migration, but what kind of people are coming—what proportion are working and generating tax and what proportion are not. In the longer run, it also depends on how long they stay. After a certain period, they will generally become eligible for welfare payments. If they stay long enough, they will ultimately be paid pensions.

Do you have the necessary data to make that assessment?

I think we do on the very latest numbers, which, if I am right, came out within the last few days—the 204,000 number, which was much lower. I should say on the back of that that our forecasts have consistently been for a very sharp drop in net migration during the course of this year. Our forecasts were that we would dip below the Office for National Statistics estimate made a little while ago about the likely longer-term number of net migrants entering the country. We already thought that we would dip sharply beneath that, but it has happened a bit more quickly than we thought. We thought it would happen during the course of next year, but it has already happened in the numbers.

You are right. That is very important for thinking about the net fiscal impact. I have not yet seen—perhaps Tom has—a breakdown of the very large number of people leaving to know how many of them were taxpayers and what their average incomes were. You need to know—

I do not think we have seen that data. That is something we would have to look at for our next forecast.

But it is possible that you will look at that for your next forecast.

I welcome the much higher headroom, but commentators point out that it is based on significant backloading, with savings of £24 billion in the final two years. Much of the tax—almost three quarters—is forecast to arrive after April 2029. We have also got the transfer of SEND provision from local government to central Government, which is £6 billion or so. Defence spending is planned to increase, and there is a real question about the level of threat versus defence spending. Against all of that—I will perhaps start with Mr Josephs to give Professor Miles a rest—how confident are you that the target of a balanced Budget or current account will be achieved in 2029-30?

There are a few points there. In terms of the target itself, the Chancellor has increased the headroom against that target quite significantly compared with the past. Some £22 billion is almost double what it was in the past. As we were discussing earlier, it is roughly equal to the average change in our forecast from event to event. It is important to underline that that is a central forecast, so the risks are in both directions—both upside and downside risks. It is also the case, as you say, that the decisions in this Budget have meant that the consolidation is more backloaded than was the case in the past. Policy decisions in this Budget increase borrowing in the short term and only decrease borrowing right at the end. Therefore the profile for the fall in borrowing and the fall in the current Budget has been pushed back. There are a number of significant risks around that, a number of which we discuss in the report. You have mentioned a couple on public spending and the commitment on defence. The Government have funded that commitment in the spending review periods through a reduction in the development budget, but the ambition to go even further beyond the spending review period and meet the 3.5% target in the 2030s is clearly a significant pressure. We estimate that reaching that target in today’s money is about £35 billion, but that is further off—I think that is 2035, from memory. You also mentioned the SEND pressure, which is very significant and more in the medium term. The Government have said that spending on special educational needs and disabilities has built up to become a very big pressure on local authority spending. The Government have said that from 2028-29, they will essentially fund all that spending from within central Government expenditure, but they have not at the current time explained how that will be funded within the current envelope, so we have flagged that as a risk.

Just reflecting on the OBR since it came into being and the fact that we have had several elections, the latest we can have an election is mid-August 2029. The year before that is when one would expect an election. All this is backloaded into years of higher political risk. Against that background, and coming back to my original question—this is perhaps for Professor Miles this time—in your experience, do Governments cut expenditure and raise taxes in election years?

I am sure they do not want to, but then again sometimes Governments do things that are unpopular, and it turns out to be quite popular. People like realism about what the Government have done.

How confident are you, Professor Miles, that these backloaded events will take place, and we will hit this headroom and this balanced current account?

It depends on how determined the Government are to stick to the plans to try to be on the right side of their target.

You are both doing very well at not giving me much by way of confidence—maybe I can ask someone else. This might be my last question. We have been talking since the OBR came into being about the forecasts of meeting a current account surplus and paying down the national debt. This is a serious issue because of the amount we pay on debt interest and resilience if we have future shocks and so on, so it is not just an academic topic. But we never actually seem to repay any debt. It seems to get bigger all the time. We talk about forecasts, but are we reaching a point where we need to move beyond forecasts? Do we just need to start seeing actual repayments and debts rather than forecasting the future?

You are right that for many years now the trajectory of Government debt has been up and up and up. One thing you know for sure is that debt relative to GDP simply cannot carry on doing that. Ultimately it will become unsustainable. In our report that we put out in the summer every year, which takes a much longer horizon, we show some pictures that in one sense are very frightening, because they say that on current tax policies and spending, if they just continue into the future, the stock of debt relative to GDP will rise inexorably and get to a level that clearly is unsustainable within a few decades. At some point there needs to be a really sustained effort not just to have the stock of debt rise and rise and rise and then just turn down ever so slightly five years down the road, but something much more significant than that. Otherwise, the thing will become unaffordable and unsustainable.

Could I just add one thing? A related point is the path of borrowing, which you asked about earlier. It is the case that since the pandemic was over, borrowing in the UK has been basically stuck at around 5% of GDP. Successive Governments have had plans, similar to the ones you will see in the report today, to have borrowing falling over the medium term, and that has not happened. Borrowing has remained about 5% of GDP, with the implications for debt that David is talking about. That has been a combination of shocks or bad news on the economy coming along, but also changes in Government plans. But you raise a very important point here about the actual position of the public finances now, rather than always focusing on what is happening in the forecast.

You state in the EFO that the Government will reduce SEND funding by £2.2 billion in ’29-30. The Department for Education have put out a statement saying the OBR is fundamentally wrong and that it will not come out of the core schools budget, and the projections do not take into account the detail of their reforms. Do you agree with the DFE and will SEND funding not in fact be cut?

A couple of things on that. One is that we have to produce our forecasts based on announced Government policy, and the Government have not announced a plan for reforming SEND at the current time. Their intention is to do that, I understand, in the new year, but there were no details that would have allowed us to score anything in this report. What they have announced is that this very large pressure—£6 billion, we estimate—will be moved from the local authority sector into the central Government sector. They did not set out how that would be funded. And that is in a year when there are other significant pressures on the existing spending review envelope, including from inflation, as I talked about earlier, but also other areas of spending pressure. It is also a year where the Government has actually decided to try to reduce overall spending a bit compared with the past. It is clearly a risk. We always have an assumption on the amount by which we think Departments will underspend their allocations. On the basis of these pressures, including SEND, we decided to take that down to zero, which is, I think, the first time we have ever done that. On the point that the Department for Education is making, we have a duty in legislation to set out the risks to the fiscal forecast. The Government had not told us in that year how they were going to fund this, or where any offset in savings would come from. Therefore, in order to show the scale of the risk and pressure it created, we illustrated the implications for mainstream school spending if—and there is an “if” there—it was all in the Department for Education’s budget.

But you scored it in the schools budget, didn’t you?

We did not make a bottom-up assumption on that, because the Government just told us that it would be absorbed within overall Government spending. They did not tell us how they would fund it. What we did was provide an illustration of the risk and the pressure—

It was only an illustration. I think the Government put something in the Red Book about it coming from across Government spending.

Yes, but we have to illustrate risk, and we chose to do it by illustrating the impact that it would have on the schools budget, which I think is reasonable.

Thank you for making that clear.

May I ask you about efficiencies in government? It seems to me that the Budget highlighted an additional £4.9 billion of efficiencies, building on what we had in the spending review, which I think was £14 billion. Can you tell us about your confidence level in achieving that? It is easy to project an extra 2% of NHS productivity, for example, but how credible is that based on what is actually attained? It often seems that people assert that to make up the numbers, but then it does not happen.

We have not built into our projections of public spending an assumption of the difference that those efficiency targets will make. Essentially, as you know, we take the overall plan that the Government set for Departments at the spending review. As I have been explaining, we make an assessment of the risk we think there is within those envelopes. On the basis of that, we adjust the underspend assumptions I am talking about, but we also try to illustrate in the book where there are risks and pressures beyond that. It is certainly the case that a number of new pressures have come into departmental spending since the spending review allocations were made, which is why we have reduced the underspend assumptions not just in the final year but across the whole spending review period. That means that, if the Government want to achieve it, these efficiency plans will need to play a really important part. As I say, we have not made a detailed assessment of those efficiency plans ourselves—

If you are talking about a 16% reduction in back office costs in the civil service in four years’ time, that is quite an ambitious number to aim for.

It is obviously the case that, in the past, Governments have been able to reduce departmental spending, most notably in the 2010s. At that time, a big part of it was efficiency savings. There is evidence that it can be done, but it is certainly always challenging.

I have many stories about how challenging that can be.

One of the risks to the fiscal forecast that you highlighted in your report is the potential for a correction in the global stock market. We are hearing news this morning that UK pension funds are starting to dump US equities, and there is a general fear that we might have an AI bubble. Could you talk us through those risks, and particularly how they might impact on the Government’s fiscal mandate?

Those risks certainly exist, and I think we did two different scenarios. In the first there was a really substantial adjustment, a 35% fall, in equity prices right across the world, including in UK stocks. In the other there was a shock emanating from somewhere else that had a smaller knock-on effect on the UK—35% in the rest of the world, but only 15% in the UK. That was because, on a lot of different measures, UK stocks just looked less overvalued than they do in a lot of other countries, particularly the US. The scale of that is very marked at the moment. One measure of valuation is what is the price of stocks relative to the earnings of companies—the so-called P/E ratio. That is not far off twice as high in the US as it is in the UK, so it seemed sensible to say that, while there could be an adjustment in the world, it is probably going to be bigger in the most overvalued markets. The UK is definitely not in the overvalued markets in the world, but you could not plausibly think that a very big sell-off in the global markets would not have any impact in the UK—so we do both of those. We do not try to identify what it is that causes this. Obviously, the thing that you highlight is very much in the minds of many people: that some tech stocks—or IT stocks, call them what you will—look like they are very stretched in terms of valuations and people are getting nervous about that. It could be that, but we are not specific. We then try to trace through what it does to the economy, but also through direct links to Government revenue, and the impact, for example, on the valuation of the stocks held by local authority pension schemes, which then has an impact on one of the fiscal measures, which is public sector net financial liabilities. A big fall in the value of stocks, given that those local authority pension schemes are big holders of stocks, is going to flow straight through to that measure—public sector net financial liabilities. The impacts are big. On the current balance, it is a deterioration by about £16 billion or so—that is as much as our estimated impact from the productivity downgrade—on the current balance, if you have 35% off equities all over the world. It is a little bit better if the rest of the world has a big hit and we have a smaller hit. It is not £16 billion off; it is £9 billion off.

Obviously, the gambling taxes were changed in the Budget. It would be interesting to know how you score those, because obviously there are an awful lot of behavioural changes. We had from the gambling industry a suggestion that people would be pushed, for example, to the black market, or equally, there is a possibility of a shift in the type of gambling, if taxes are put up on some. How do you go about scoring that and what were any particular messages about that policy?

This was quite a complicated measure to cost, and there are a number of assumptions in there on behaviours. The changes create quite a number of different rates across different forms.

It is quite complicated.

It is. There are different rates across different forms of gambling, so we also had to consider the impact of that on behaviour. Essentially, the main behaviours that we assumed, based on studies that HMRC has done in the past and updated looking at international examples of where gambling tax has changed recently, of which there are quite a few, I guess the big one is that operators will pass on a lot of the cost of the tax increase to their customers. That would lead to a reduction in demand, which reduces the yield from the measure by a reasonable amount. That does build in some behaviours of people moving into the illicit market. The evidence we had from those other countries that have introduced reforms recently is that that they did not see huge evidence that that has happened. Especially if you have a well-regulated gambling market, as we do here, the opportunities to do that are relatively limited, and generally people do not want to do that if they can, because of the obvious risks involved. We did also have an assumption that there would be some substitution between the different forms of gambling, reflecting the fact that you do now have some quite wide differentials between the tax applied on what are essentially quite closely related forms of gambling. We also made an assumption that the operators would, over time, be able to adjust their product offering to essentially maximise profits and minimise the tax burden.

So from what you are saying, on online gambling—online gaming, which this Committee reported on, and which we felt caused more harm, hence we urged the Government to look at differentiation—you factored in that they might try to approach some of those people to go, for example, to a betting shop or the races or a casino. Is that realistic?

I should say that it is a relatively small offence, but we think it is likely over time that some of that behaviour would take place. Again, there is a big amount of uncertainty about that behavioural response, but it is one where HMRC have done really detailed work, using their own analysis from the past and taking account of international evidence.

We will perhaps raise it with them when they are in front of us in January. Thank you, Mr Josephs.

There were new taxes to electric vehicles applied in the Budget. Could you set out what the scope of this tax is to plug the fiscal gap created by reduction in fuel duty revenue?

We have some analysis of that in the report. We have for a long time flagged in our long-term risks report the big risk to Government revenue from the loss of fuel duty over time, as the Government shift to meet their climate change objectives through electrification. We did some analysis of the offsets that the new EV charge will create, and what it showed was that essentially it offsets around a quarter of the lost revenue from fuel duty over the long term. It is not 100% because the EV charge that the Government have set is roughly half the charge on the average driver through fuel duty. Also, currently it applies only to cars, not to vans and lorries, which account for close to half of all fuel duty receipts. Those two factors—the lower rate and it applying only to cars—are why it replaces only around a quarter of the revenue from fuel duty.

Would your expectation be that the charge will have to increase in the future to fill that gap, and potentially be extended to vans and lorries?

That is obviously a decision for future Governments, not for us.

You predict a reduction in EV car sales of nearly 450,000 as a result of the measure, which is offset by other Government policies to increase sales. Could that reduction in sales be bigger than envisaged? How confident are you around the modelling of the impact of the policy change?

Again, it is quite an uncertain behavioural response, and there are quite a lot of moving parts. There is the actual new charge on electric vehicles, and the impact that has on purchasing decisions, there are the offsetting policies that the Government have introduced to try to incentivise purchases through grants and other changes to taxation, which will reduce costs, and there is also the interaction with the electric vehicle mandate—manufacturers have to sell a rising proportion of electric vehicles. Essentially, the interaction of all those moving parts leads us to the judgment that overall, on a net basis, sale of EVs will be about 100,000 lower than they otherwise would be. But that is very dependent on assumptions you make on the responses of both the manufacturers and customers.

So there are significant risks around the figure?

Yes, there are definitely risks around it.

On fuel duty, the Budget states that from April 2027 the Government will uprate fuel duty by RPI. But as we know, fuel duty rates have been frozen for 15 years. How much headroom would the Government lose if they did not go ahead with uprating fuel duty by RPI?

I think this is now the 16th consecutive time that those rates have not changed. We estimate that the cumulative cost of that since 2010 is about £120 billion. In this forecast, if those rates were not uprated, it would cost about £3.5 billion a year.

Say that fuel duty was not uprated: does that potentially have an impact on demand for EVs, and do you model that?

It would do. Essentially it lowers the cost of petrol and diesel vehicles relative to EVs. The modelling that I described earlier, on electric vehicles, we have to base on current Government policy, so it will have been based on the assumption that those RPI increases happen.

So that number on the EV forecast could change, in the negative, if Government were to freeze fuel duty?

Yes, but as I say, there is also the interaction with the mandates. Manufacturers still have to sell a certain proportion of their vehicles as electric vehicles. So again, there would be a response to that from manufacturers.

In the last Budget, there are several increases in the effective rate of tax on income flowing from wealth and, in particular, in the effective rate of tax on dividends. I see that, in relation to behavioural effect, you find that, over the five-year forecast period, the effect of people changing their behaviour in response to that tax increase goes down to zero. Was that a surprising find, and would you have found a zero impact on behaviour if that increase had been larger than 2 percentage points?

Specifically on the increase on dividends?

Yes.

I think that reflects the fact that it was a relatively small increase. It applies just to basic and higher rate taxpayers and not to additional rate taxpayers. I think that quite a lot of the dividend taxes will be paid by additional rate taxpayers, so that is essentially why there was not a huge amount—

No, but can I just say why?

No, in the sense that we did not judge that the positive effects outweigh the negative effects. I think it is in some ways not surprising that the answer is no, because if you are going to take the tax take out of GDP higher again, which happens to the highest level certainly since the second world war and possibly for a much longer period—it is very difficult to raise taxes without affecting incentives to save, to work, to invest, to become an entrepreneur and to stay in the UK. It is very difficult to do that without having some negative effects. We judged that those negative effects, at least over a five-year horizon, were not large enough actually to give a negative impact from the Budget, but neither did any positive effects more than outweigh them.

It is very difficult to increase the tax take without having effects on incentives.

Well, again—

Well, I think it is difficult. We perhaps need to be a little more reflective than that.

It is the same answer that David just gave, which is no, but there is a reason why, which is that we have introduced the new significance threshold, so we incorporate the impacts of policies on productivity only if they go above or below our 0.1 threshold, and in this Budget, there were not any policies in either direction.

I am afraid I do not have an answer to that, because that is not a number we have included in the Book this time. I know there have been some recent data releases on that, but I am sorry: I cannot give you a number.

I need to chip in that my local hospital, Homerton, has increased productivity by 11.3% year on year, so there are bits of the NHS that are doing very well.

Yes.

As you said, we have not changed that assumption because we do not feel that there is any hard evidence that would lead us to change it. That assumption is that there is pretty significant outward migration as a result of that policy.

Yes, there is no new evidence, and we will not have hard data on that until early 2027, when we start to see the actual tax receipts.

You would expect it to have an impact over the long term. It reduces the tax benefits of the salary-sacrifice route to pension saving, and over the long term, therefore, you may expect lower pension savings than you otherwise would have done.

There are obviously lots of other routes through which people can get the tax benefits of saving into a pension. The costing assumes that people will move into other routes.

I have a question about the significance threshold on policies. Do you only ever assess what the Government put in front of you, or do you have a basket of measures, as it were, or a range of different types of policy, that you publish, and that you could show make the biggest difference in terms of enhancing growth? How does that process work?

We only consider policies that the Government have announced as final policy decisions. That is set by legislation. We are not permitted to look at alternative policies. One of the issues that we have faced when undertaking these sorts of exercises, incorporating the impact of Government policy, is: how can we be assured that we are being given a full picture of the various changes in Government policy that might have an impact on growth? In particular, that is an issue for us when it comes to departmental spending policy and regulatory policy. On tax and welfare rate changes, essentially, fiscal events bring everything together, so we can be very confident that we are looking at all tax and benefit rate changes. When it comes to departmental spending, for example, we might be told by the Government that a Department is introducing a new spending policy that will have a positive impact on growth. It is very difficult for us to look through everything else that Department is doing and judge whether other things it might be doing may be having an offsetting effect on growth. That is one of the reasons why we have introduced this significance threshold. We have also been clearer on another one of our thresholds, which we call additionality, which essentially asks, “Can we be sure that this new policy is having an impact on growth, which is clearly over and above previous Government efforts and is not being offset by other policy changes?”.

I appreciate that you only assess the policies that the Government put in front of you, but to ask the question slightly differently: do you have a typology of policy measures that shows how you assess different types of policies in-house?

Yes, we do.

Is that something that you publish?

We do, and the published report shows how we approach different types of policy.

The Employment Rights Bill was in ping-pong immediately before the Budget, but I suppose the U-turn on Wednesday, on first day rights, would vindicate your position on that. Is that right?

Back in March, we said that the final picture on that policy was very uncertain, because it had to go through a legislative process with a lot of amendments in play and quite a lot of the important detail only coming in secondary legislation. We thought that by the time of this forecast, we would have a clearer view but, as it turned out, that was not the case—it was still going through this process—so we judged that we were still not able to estimate the impacts on growth and, as you say, there have been subsequent big changes.

It is clear in legislation: you can only deal with the policies of the Government. It is the same for legislation—it has to be finally pinned down.

We do score policies where the legislation is still going through Parliament, when it is in most cases pretty clear what the final outcome will be. In this case, that was just not the case.

Because of the number of amendments and—

The number of amendments, and the fact that a lot of important details would come from the secondary legislation that followed the primary.

That is a message for Government, isn’t it, about what to put on the face of a Bill and in secondary legislation? In government, you always want it in secondary legislation; in opposition, you always want it on the face of the Bill. Before we finish, I have a couple of quick questions. We have not talked much about gilts today, but we have seen the pattern of gilt trading. One big question is, why are UK gilt rates consistently higher than those in France, which has far more debt than even we do?

Part of the reason is that UK gilts are denominated in sterling and French Government bonds in euros, and the euro is a clearly a more widely held global currency than sterling. A lot of investors will think some part of their portfolio needs to be in euro-denominated assets and relatively fewer, at least outside the UK, think, “Yes, we absolutely have to have some sterling-denominated Government bonds”. That is one reason. A slightly more nebulous one is that, if things got really fiscally difficult in France, there might be some European Union or ECB strategy to help them. That will not come our way.

That is very helpful. Going back to what we were discussing at the beginning, you were clear about the unhelpful information that you felt was being put out there. If that unhelpful briefing occurs for a future fiscal event, it is clear that the Committee would like to be informed. Are you happy to support that, as members of the Budget Responsibility Committee? Obviously, you will have a new chair who has to be involved in the discussion, too. Would you support that, Mr Josephs—providing information if you felt that the OBR was being buffeted by briefings?

Yes, absolutely. With actual briefing, it is often difficult for us to know exactly what is happening, who is briefing and what.

It could be wild speculation, media speculation, briefings, leaks or all of those—the noise around it, I think, is my point.

One thing I might add to the discussion we had earlier, when you were asking about the letter, is what was uppermost in our minds when we considered writing that letter: the expectation that this Committee would ask us questions about what had happened, because you have done that previously. That was therefore part of the thinking behind putting it down on paper, so it was clear.

Get it out early.

And we had had questions from the shadow Chancellor.

What about you, Professor Miles? Would you support providing us with that information in future?

I would. The OBR is here to serve this Committee, and if you find that information helpful, we should respond.

Thank you. Did you ever feel—I will push you a little more on this—that the Chancellor or any Treasury Minister gave an incorrect impression to journalists or the public about the information that you had supplied to them?

I think there were misunderstandings in the media, but that is that is not something that I would pin on—

You cannot be sure where it came from.

I feel the same. It is very difficult for us—we are not here to comment on Government communication or on what then led to media speculation. The purpose of the letter, as we have discussed, was to set out the facts in terms of the evolution of our forecast. Everything else that happened is not really something for us to reach a judgment on.

Thank you. I thank our witnesses very much, Tom Josephs and Professor David Miles of the Budget Responsibility Committee at the Office for Budget Responsibility. This is the first of four sessions on the Budget: tomorrow we have a group of economists; next Tuesday we have the Bank of England; and on Wednesday we have the Chancellor of the Exchequer. The transcript of this session will be available on the website, uncorrected, in the next couple of days. Once again, I thank our witnesses.