Sir Robert Chote and Andy King.

Q43 The Chair: Welcome to the Lords Economic Affairs Committee’s fourth evidence session in our inquiry into the United Kingdom’s fiscal framework. We are delighted to have with us Sir Robert Chote, former chair of the Office for Budget Responsibility, and Andy King, former member of the Budget Responsibility Committee. Thank you both so much for coming. We really appreciate it. The session is being broadcast live on Parliament Live TV and we will give you a full transcript shortly after the session to make any corrections.

I will start with a general opener. What is your sense of what we can say about the effect of the OBR over the time that it has existed? How do we know whether the OBR has been successful? What is our metric for it? In addition to that, could we conceive of a fiscal council as being effective without fiscal rules alongside it? Is there a role for a fiscal council independent of the interaction and monitoring of fiscal rules? Robert, perhaps I could start with you.

Sir Robert Chote: Certainly. Thank you very much indeed. It is a great pleasure to be here. Yes, I think we can see a value in fiscal councils and in the OBR in the absence of rules. That is primarily about bringing much greater transparency across the whole piece of the public finances, the way in which we assess the current state of the public finances, how we expect they are most likely to evolve under current policy and, crucially, what uncertainties and risks lie around whatever central view we may take.

Obviously, there is a lot of focus on the forecast and how the forecast interacts with the fiscal rules, but one should not underestimate the value of that transparency: the weeks-long progress of the OBR interrogating and challenging officials from the Treasury, HMRC, DWP, the Debt Management Office, you name it, on all the innards of the public finances, and surfacing that for the benefit of Parliament and the public.

There is then the value of the forecast that it produces, and here we need to distinguish a bit between the OBR and fiscal councils in general. The OBR is relatively unusual in producing the official forecast for the economy and the public finances and incorporating the policy measures. In many other countries, the role of the fiscal council is either to comment on the reasonableness of the forecasts and the interaction with whatever supranational or national fiscal rules there may be, or indeed to produce a parallel forecast alongside which we can compare it. It works in different ways in different countries, reflecting, typically, where power resides in budget setting in different countries.

On policy outcomes, we cannot run the counterfactual of what the debt-to-GDP ratio today would look like if the Treasury or the Chancellor had continued to be responsible for the forecast rather than the OBR. I would like to think that at the margin, the OBR has encouraged better policy decisions than there otherwise would have been. I have to put my hand up and say that the public finances were worse when I left the OBR than when I arrived, but I do not think that was entirely down to me—some external shocks came along.

I think the real value is in having the transparency and integrity of independent analysis, with which reasonable people may disagree, and putting that forward for the benefit of the public, investors, voters and Parliament.

The Chair: That is very interesting. Thank you very much. Andy?

Andy King: Thank you for having me here. There is another way of thinking about this. I know you have already spoken to a number of academic experts and think tanks, but the evidence is that fiscal councils tend to work better in tandem with rules. The way you can think about it is that the role of fiscal rules is to constrain the discretion of the Chancellor and tie her hands when she wants to spend more for short-term reasons, and the fiscal council, alongside the fiscal rules, creates great transparency about what you could call the fiscal reaction function.

The world is a volatile place. The outlook for the public finances will be changing all the time. The combination of fiscal rules, which help guide where that volatility goes and how we offset a negative or a positive piece of news for the medium-term public finances, and a fiscal council that is setting out the numbers in the way the OBR does, allows everyone to see very clearly that when X happens, the Chancellor responds with Y.

At the OBR, we were able not just to do the numbers but because we told our own version of the story—the numbers purely, not the politics or the motivations—we could say, “Presented with the following hole or the following surplus, the Chancellor chose to do X, Y, and Z”, which sets out nice and clearly that reaction function. That has always been important and something that Chancellors think about, but it is much clearer when there is a fiscal council telling that dry, numbers-based story alongside the Budget story.

The Chair: Thank you very much, both of you.

Q44 Lord Verjee: How have fiscal rules and watchdogs changed the role that bond markets play in enforcing fiscal discipline? Has the impact been entirely positive? As a follow-up, to what extent do markets care about Governments meeting fiscal rules, and what other considerations do they weigh when deciding whether to lend?

Sir Robert Chote: The key value in what has changed with the bond markets is that fiscal councils and the OBR provide a greater, richer and more robust information set for people to make decisions on whether and on what terms to borrow or to lend to the Government, and that has to be a good thing.

On what effects it has, if I knew precisely what affected bond yields, I would not be sitting here, I would be on a yacht. However, as I say, the transparency that it provides around the policies of the current Government is clearly an important piece of information for investors to take into account. I suspect that if you are deciding whether to invest in UK equities or UK bonds over a longer period, you will take a wider view not just of what this Government are saying they will do but what they might do, and what other potential Governments might do as well. Better and more robust information as a backdrop is likely to mean that decisions are better informed, but that other factors will be taken into account.

Obviously, one area that you do inject is that the nature of the Government’s commitment to and relationship with the fiscal council, and their commitment to fiscal sustainability and sensible policy will be more in the light as well. There may be a situation in which the Government are or are not taking the fiscal council seriously or dealing well with it and that may, in addition, affect the view of a sensible rate at which to lend to the Government.

Overall, the value, as we discussed before, for investors, as with voters and parliamentarians, is greater transparency based on professional judgment, not on wishful thinking.

Andy King: The only thing I will add is that the bond market is still there as the ultimate backstop that will punish Governments who do something that is not sensible, as we learned three years ago. That is still there. In some ways, fiscal rules and fiscal watchdogs are there to help Governments not get too close to that ultimate backstop. You do not want a Government to be running too close to the bond market, punishing them and having all that information, and rules-based policy setting is one way of avoiding that.

Sir Robert Chote: The idea that the bond market is obsessed by whether you meet or miss a particular fiscal rule in a particular year by £1 billion one way or the other is not very likely.

Lord Petitgas: I agree with you that information is the red cells that go into the blood of the market, and you have been on top of this for a long time. Given the fact that we have some fiscal tightening and we have a very robust OBR structure, how do we explain the gilt markets putting such a big premium on the rate we pay as a country versus others in the G7, including those in Europe who I think are in a worse position?

Sir Robert Chote: As I say, if I knew the answer to that question, I would be a lot wealthier than I am. There is obviously a number of factors. There are clearly developments in the UK bond market that are fundamentally a reflection of global trends. We have seen the responses, for example, to whether US trade policy is looking more or less inflationary at any given moment. There may be other structural factors around the demand for government debt that may be reflective of regulatory or other issues. As I say, I think trying to tie very closely a short-term movement in the bond market to some particular development in a forecast or a rumoured development in a forecast is probably overengineering it somewhat.

My bigger concern, stepping back, about the confidence the markets and others should have in fiscal management in the UK is: do we believe that we are in a situation where not just this but other Governments are willing and able to take difficult decisions and get them to stick in Parliament and that we will not simply see the debt-to-GDP ratio and servicing costs ratcheting higher and higher, shock by shock, because we never have the period of consolidation and the rebuilding of fiscal space between one negative shock and the next?

Q45 Baroness Wolf of Dulwich: My questions follow on rather neatly from what you just said about periods of consolidation. We have talked about fiscal rules and watchdogs in general, but I would like to ask some questions about the specific sort of fiscal rules that we have in this country. Do you feel—particularly given the fixed-year timeframes that get stipulated—that they stop us thinking about the economic cycle? Our focus is year by year by year, and that may quite often mean that we get a countercyclical response. Even more specifically, does the way that they are framed, as someone has suggested, incentivise Governments always to operate without a sufficient buffer?

Sir Robert Chote: There are several elements in that. We do not have a rule for a cyclically adjusted measure of the deficit, which would obviously be one way to bring that into play but would have all the complexities of how we estimate the current size of the output gap, and the relationship between the output gap and the structural deficit.

Baroness Wolf of Dulwich: I was not suggesting another super-specific rule. It is more: does the way that we look at them in fact mean that we just do not think very sensibly across the cycle?

Sir Robert Chote: If we look at the five years and the target is in year 4 or in year 5, the Government can focus very much on whether the letter of the rule is met in that particular year. I think markets and others would base their confidence or their appreciation of the fiscal strategy based on the period as a whole.

We have, in this most recent Budget, yet another Augustinian fiscal package of “virtue but not yet”, a period in which measures are adding to government borrowing in the near term, then tightening quite significantly at the point at which the target is biting—albeit the target will, of course, move forward to a three-year from a five-year horizon and there is a bit of heading off for that. None the less, you have to worry that this year’s year 5 will be year 1 a few years down the line, and at that point you are more likely to be giving away than not. The OBR has pointed out endlessly the tendency to these Augustinian policy packages.

Sufficient room is really a judgment for the Government at the time as to what risk they want to take that the rule will be breached. The OBR produces a median forecast so, by definition, you would think that there is a 50% chance it will be higher and a 50% chance that it will be lower. If you want your odds to be 50:50, fine, set policies such that the target is absolutely met, but if you want to leave yourself with a greater than 50% probability that you will meet it, you need to build in some room for manoeuvre. That is not costless because that money could be spent or used elsewhere, but if you want a credible rule structure that is not prone to policy settings being as volatile as the forecast inevitably will be—and that is not a failing; that is a feature because of market changes, data changes, judgment changes and so on—it is sensible to put in some room for manoeuvre so that you are not having to move policy around in response to every movement in the forecast.

The other issue with fiscal rules—and again it comes back to my worry about things ratcheting up over time—is that if you are in an environment in which when you get a pleasant surprise, you spend it, and when you get a bad surprise, you change the rules or push the time horizon out, that contributes to an unhelpful ratcheting over time.

Andy King: Of course, I agree with all this. When I look at the fiscal targets as they are now, the current balance target seems to me a very conventional—at least for the UK—golden rule type of target, whereas the balance sheet target for debt to fall, on a very broad measure of debt, is about as loose as it is possible to be. It is not totally consistent with fiscal sustainability because you can have debt rising for ever and meet that target. Therefore, in an objective sense, the fiscal rules are reasonably loose. You can make a perfectly reasonable case for doing that because it allows you to address a long-run public underinvestment problem. The argument would be that these are not normal times, therefore we should allow debt to rise for public investment.

However, the lack of fiscal headroom, or the running with a tiny buffer, seems to me entirely a political construct. It is like any fiscal buffer is not treated as a buffer. It is a war chest. It is available to spend. The Chancellor has to fight to even keep what are, in macroeconomic terms, tiny amounts of headroom, and if it gets too big that is a problem because all the bids for spending it will come in.

In a sense, the broad structure of the fiscal targets as something that will ultimately be three years ahead, with an escape clause for major shocks, seems to me like the right thing to have. The specific targets are reasonably tight on the flow but reasonably loose on the stock. However, the fiscal reaction function that you want to run with tiny amounts of headroom means that, as Robert has just said, all the volatility that is in the real world, and therefore gets reflected in the forecast, is finally reflected in policy changing frequently, rather than headroom going up and down as the world changes. That feels like the really suboptimal thing, and it is the motivation for wanting a single fiscal event, trying to downplay the headroom assessment in the spring, and so on. People are very aware of it, but it is a problem within the UK fiscal architecture.

Lord Blackwell: To come back on the alternative to a balance sheet target, the problem, as you both said, about a one-year target for spending versus revenue is that it is the difference between two very large numbers and is, therefore, inevitably, hugely volatile. It is a bit like setting a course now to land an aircraft on an aircraft carrier five years out—there are so many things that could change.

However, the objective, why we are doing it, is to control the level of public sector borrowing and the level of public sector debt. If you had a balance sheet target, which might be the same or might be lower, that five years out public debt should be no more than X% of GDP, it has the merit that it is cumulative so you cannot just fiddle the last year, it has the merit that it is a large number and therefore small volatility would not change it very much, and it is the real objective. Why not say that we want to have public debt as a percentage of GDP X% five years out or less than X% five years out? The OBR can say, within 80%, “Here is a lower and upper range of whether you are not going to hit that”.

Andy King: My experience of fiscal targets, of which we have had quite a few over the last few years, is that all of them have their challenges. It may feel attractive to say, “Why are we targeting the change in debt when what we care about is the level?” If you have a level target, any shock in the near term means that to get back to your level target, you have to introduce very countercyclical policies. It fails the requirement to do the sensible thing in some situations. However, the target we have at the moment allows us to ratchet constantly higher, which is also bad. There are downsides to both types of rules.

Having looked at many of them and thought about what they deliver in the longer term, I feel like the idea of having a target that boils down to the sensible idea that in normal times debt should be falling is the right thing to do. The way it is set up at the moment is okay.

The problem is that ratcheting up, “We will borrow more in the near term and we will tighten in the target year or tighten for the target year”, the problem of headroom, but also the problem of not taking into account the fact that the world will deliver shocks. Forecasts do not contain shocks because you do not know in what year they will happen and you do not know the nature of them. You describe it as a risk. However, you can set targets in the knowledge that shocks will come along.

The OBR did some analysis on this, either just before I left or just after I left, looking at, “What is a reasonable number? How much should debt be falling in normal times if you want debt not to rise after the effects of shocks?” I think the number was 1% to 2% of GDP. If you think that we are currently targeting it being minus something very small, the change in debt in the target year, and if you think that maybe it should be minus 1% or minus 2% to have debt not rise in outturn, that is a very big fiscal tightening. It is a huge political choice. It may look right from the perspective of a technocrat designing a fiscal target, but you have the combination of starting from here with debt very high because of three massive shocks and the attempts to deal with them, and all the pressures going forward of an ageing population and so on. It is incredibly difficult to get from here to where might be ideal.

The Chair: Your comments are so interesting that we have a couple more follow-ups.

Lord Agnew of Oulton: I want to understand that in simple language. You are saying 1% to 2% of GDP in normal times. That is, in current money, £25 billion to £50 billion a year going directly to debt reduction. Is that broadly what you are saying?

Andy King: As a share of GDP. You are not paying off that much debt, but you want debt declining.

Lord Agnew of Oulton: Yes. You are holding it, though, at that sort of level?

Andy King: In the simplest language, it is because, say, once every 10 years, you will get a shock that pushes you up to 10% of GDP. In normal times, for the rest of the decade, you want to be going down one at a time, so that when you go back up, you are back to where you started.

Sir Robert Chote: This is a particular challenge if you have a shock that is increasing the structural deficit. Then you are turning the tap off in the bath, but the level of the bath water continues to rise for quite some time. The question is how aggressively you should go at that to ensure that you have the fiscal space to deal with future shocks. It is a case that has to be made, saying that this is an investment in the future, in the way in which some other investments are as well, for future benefit.

Lord Burns: To some extent, this answers the question about why the OBR has not been more successful over the last 15 years in preventing the debt ratio rising. One of the features of this has been that in scarcely any of the years since then has the debt ratio come down. What you are saying is that this is not a problem with the OBR; it is a problem with the design of the rules, which have encouraged a system whereby when there is slack and there is margin, you spend up to it, and in tough times, as you have said, you revise the rules, whereas to get the debt ratio down in the good years so that you are ready for the bad years will require quite significant tightening of fiscal policy relative to what people have been conditioned to.

Sir Robert Chote: That is right. The role of the OBR is to make those points and to go through the sort of calculations that Andy has described. As you know very well, ultimately it is elected Ministers who decide the rules and how to respond to shocks given those rules. Do you change them? Do you change the policy? Do you say, “I am happy not to meet these for the time being because I want to see how things evolve or what other shocks may come along”?

The OBR is informing these decisions, informing Parliament and everybody else about the choices that the Government have made, and indeed, usefully, looking back over time. There is a pattern here where, for example, the Government say, “In future years, we will raise fuel duty in line with inflation”, and they never do. There is this tendency to an Augustinian pattern in packages. It is an important part of the role of the OBR to point that out. At the end of the day, it is for Ministers—

Lord Burns: Could it point it out more forcefully, do you think? In a sense, as one reads the OBR reports, they are often just playing by the rules of the game rather than pointing out that maybe this rule is not going to get us to where we want to be.

Sir Robert Chote: Fulfilling the requirements placed upon it by Parliament, I think is what you mean. It is an occasional frustration that the EFOs get an awful lot more attention than the fiscal sustainability and risk reports, which go into this in a good deal more detail and where some quite innovative work has been done, not least with Andy’s work on net zero. The more that other entities can bring attention to that work, the better.

Andy King: The OBR’s job is to shine a light on lots of things. Running tighter fiscal policy in the aftermath of the financial crisis is not obviously something that would have been better for the country. Now, with the more recent Augustinian stuff, possibly you can make that case. One of the things we shone a light on in the mid-2000s was the extent to which spending on public services was falling as a share of GDP to levels not seen for decades, despite the much greater calls on public services in an ageing society and so on. In all these cases there are massive trade-offs. If you are trying to right the fiscal ship after a big storm at the same time as dealing with rising pressures on public services from all sorts of quarters, and public investment has been low and therefore the quality of public assets is almost literally crumbling in many cases, it is not an easy set of trade-offs.

Lord Blackwell: The formulation of a target of debt as a per cent of GDP five years out, with the OBR then saying what the upper and lower probabilities were around that, would allow the OBR to factor in the risk of a shock. You could, in that sense, encourage fiscal conservatism by pointing out that the course had a low probability.

Sir Robert Chote: We have, throughout the history of the OBR, put probabilities as best we can around the particular fiscal rules of the day being met. We did that for many years based simply on the size and distribution of past forecast errors. It got more sophisticated when I left, probably because I left, with Monte Carlo simulations and so on. We inevitably used more scenario analysis in the wake of the Covid shock, for example, because talking then about what the uncertainties were around a central path, looking at the nature of the forecast differences in normal times, was not very informative, and we very sensibly pivoted to that sort of scenario analysis as well.

All the discussion about the right level of fiscal room for manoeuvre and headroom against the target, to which there is no correct answer, is informed by a very nice history of the sizes of the margins for error and the probabilities implied by those for the likelihood of hitting the target on current policy. Right from the start, we wanted to be clear about the impact of uncertainty. No policymaker should bet the farm on any forecast, theirs or anybody else’s, being right in every degree, because it will not be.

Q46 Lord Lamont of Lerwick: Many commentators have described the OBR as policing government policy. Is that a fair assessment or a fair description? Is government—talking about Governments rather than one Government—sufficiently clear that its scope for action against the background of a fiscal regime or fiscal rules is self-imposed, or is it too willing to accept a framing in which the OBR constrains its scope for action?

Sir Robert Chote: You are right: there are occasions on which Governments, in essence, say that they have been forced into a decision by the OBR or by some forecast change that the OBR has made. That is nonsense. The OBR is providing its best assessment of the outlook for the public finances, including variables that the Government may have decided they want to target on the basis of current policy, and it talks about the uncertainties around those.

If the OBR comes to the Chancellor in the weeks running up to a Budget and says to him or her that they are not on course to achieve the target that they have set themselves, it is for them to decide. Do they change the policy so that when the published forecast comes out, they are hitting it? Do they change the rules? Or do they say, “Actually, we are content for this forecast to show that we are not hitting the rules, but we will come back and address that later”, or “We want to wait and see how things turn out”? All those things have been done by different Governments over the period.

Occasionally, as I say, it is framed very unhelpfully as the OBR forcing decisions. The OBR does not do that. Governments make decisions based on their own targets.

Lord Lamont of Lerwick: Can you give some examples of where the rules have been changed in the short term in reaction to a view expressed by the OBR?

Sir Robert Chote: They have been changed not so much in light of the view expressed but the knowledge that you will miss the rules if you do not change the targets. Essentially, you have had either concrete policy decisions such as the increase in NHS spending in 2017, for example, which came out of a Budget cycle, or specific forecast shocks that force you to move to longer horizons. I think we have been through about eight to 10 variants of the rules over time, and it has been mostly for those reasons.

Andy King: Often, it is the OBR delivering the message that the pandemic has had a bad impact on the public finances, but everyone would have delivered the same message. After something as extreme as that, it probably is right to change the rules, as was the case after the financial crisis.

There are examples as well. The rules were changed in the aftermath of the Brexit vote, or in fact just before the Brexit vote. They had been tightened up a lot after the 2015 election, to the point where they were being gamed quite significantly to deliver a surplus. They were changed after that. They were also changed more recently as Chancellors changed, and so we got to the stage where they were being changed every couple of years, were they not? Changing the rule was one of the go-to measures for dealing with any fiscal problem, and the rules were, for the most part, loosened. That was the other feature that I am sure Lord Burns did not approve of.

Lord Lamont of Lerwick: Could I ask about the scoring of growth measures? Daniel Susskind, who I am sure you are familiar with, wrote not so long ago that “the idea that the OBR somehow knows enough to take each UK government policy and state its impact on growth to a single decimal point is fanciful”. Obviously in any Budget, there is a whole lot of measures, tens of measures perhaps, that will contribute to growth, but they are all small and individual. Is the system designed to take account of this sufficiently or could there be a bias towards fiscal orthodoxy and fiscal conservatism?

Sir Robert Chote: I do not think so. You are producing your best assessment of what the outlook is over the five-year period, based on the policies that are there.

When people talk about the way in which the OBR takes account of the impact of policies on growth, you need to make a distinction on the demand side and the supply side to start with. Clearly, at one level, there is the question of—which tends to apply to the package as a whole rather than to individual measures—is this essentially a giveaway or a takeaway Budget that, all other things being equal, will increase demand in the economy and increase activity? You are then confronted with the question of: what is the Bank of England going to do with monetary policy in response to that, given the desire to hit the inflation target? A big giveaway Budget might mean more growth in the near term and less growth in the longer term, as you get nominal GDP back to roughly what the Bank of England thinks is consistent with the inflation target.

There is a separate question of the impact of policy measures and policy packages on the supply side, on potential GDP in the future. The OBR produced a paper on Budget Day this time looking at, I think, 19 occasions on which it has changed estimates of potential GDP reflecting policy measures that have been announced in Budgets between 2023 and 2025. As I say, there have been about 19 occasions on which the changes have been large enough for the OBR to deem them significant enough to include. The majority of those were because of their impacts on labour supply. I think one was a planning reform that it concluded would affect potential GDP via total factor productivity.

You are absolutely right to point out that a lot of these sorts of measures have relatively modest effects, certainly over the early part and through a five-year forecast horizon. They might have larger effects beyond, which the OBR typically flags. Changes in the public sector capital stock you might expect to be beneficial for potential GDP, but that would be much more apparent after 10 or 15 years than after the five over which the forecast takes place.

All this is against the backdrop, of course, of potential GDP being impossible to observe directly and very hard to estimate. Certainly, in my day, I had quite a high bar for making small, precise changes to an estimate of potential GDP that you thought had a huge range of uncertainty around it in the first place. More recently, for perfectly understandable reasons, the OBR has been more transparent about the way in which it includes those things, perhaps at the price of, in the run-ups to Budgets and fiscal events, Treasury officials being urged to come to the OBR with endless streams of paper suggesting that these policies will be transformative for growth, in a rather unconvincing way.

Lord Lamont of Lerwick: Presumably, on supply-side measures, there could be a wide variety of opinions. There would not necessarily be a right opinion. There might be a conventional one, but not necessarily a right one.

Sir Robert Chote: Absolutely, and that would be true, of course, of the forecast as a whole. I do not need to remind you of the panel of independent forecasters, a temperamentally and ideologically diverse group where I think you could have put your forecast within a very wide range of “reasonable views”.

Andy King: For the growth side, the supply-side scoring or dynamic scoring as they call it in the US, basically the process has been changed probably twice. For a long period, there was quite a high bar for any supply-side measure being factored into the forecast explicitly. The one I remember most clearly was the national living wage being brought in at a much higher percentage of median earnings, and us feeling that we had to look at that in isolation and think through what it would do to the economy. Then, since 2022, the bar was lowered to be more active in taking supply-side impacts into account, and more transparent—there were a lot of reasons for it. In this Budget, that bar was raised again and it was made explicit, so only measures that have an impact on GDP of more than 0.1 in the final year of the forecast will be scored.

When I think about these things, and when I used to be responsible for thinking about these things, there is analytical purity and there is political economy. On the analytical side, it is obviously better to think about the supply-side impacts of tax measures and regulatory measures that change incentives in the economy. If you sit down and look at that thing on its own, you will say it is good for growth or it is bad for growth in a medium-term, supply-side sense. However, the political economy of it is very difficult for the OBR-Treasury relationship. When the bar was lowered, the Treasury was very fast to identify the pieces of evidence that the OBR had said were sufficiently robust for things to score, and all of a sudden there would be a long list of measures that use that same academic paper, those same elasticities and so on. Also, more generally, it would go out to departments and say, “Do you have supply-side measures? We can now throw these into the mix”.

The quote you started with in a sense is correct, but sometimes the evidence on the marginal impact of changing, say, an income tax rate is quite good. The underlying forecast is hugely uncertain, but the fact that you will be a little bit better than something if you do something to a tax rate is there.

I do not think that the more general critique of the OBR that because it is so fiscally conservative and does not take into account supply-side measures, it is understating growth prospects, stacks up in reality. There was a Budget, when Jeremy Hunt was Chancellor, where I think he touted that he had announced the 100 growth measures. I guess they are taking effect right now, but there is no sign that the OBR was massively too pessimistic about that growth package in the data today.

In a sense, they are borne out in the general argument that moving the supply side of the economy is difficult because it is very, very large stocks of capital, of employment and so on, and things tend to move slowly. However, that political economy side has to be taken seriously. Drawing a very firm line in the sand at 0.1 creates a boundary that the Treasury can now look at. We will try to find positive measures that tip the rounding and will be taken into account, and we will try to find tax measures to pay for it that have a supply-side impact just below 0.1 so that they will not score. That cliff edge in the OBR scoring is the next thing for the Treasury Committee in the other place to look at, but wherever there is a line—you see this with fiscal targets as well—there is an incentive to push the boundaries. This is a new line to look at.

Sir Robert Chote: We will presumably also see measures artfully combined together so that they are a single measure that gets over 0.1.

Q47 Lord Razzall: Just finishing off on that, it sounds to me that the two of you are getting very close to JK Galbraith’s quote about economic forecasting, that it is there to make astrology look respectable. That is not my question, by the way.

I want to look at the international comparisons. Did you in your previous lives ever look at other fiscal oversight bodies in other countries to learn from their approaches? If so, is there anything that we can learn from that?

Sir Robert Chote: The answer is yes. We were very actively involved in the OECD’s network of independent fiscal institutions and parliamentary budget offices, which are—

Lord Razzall: Five of them?

Sir Robert Chote: Fifty. The IMF has a regularly updated database of these bodies and there are now 55 of them in, I think, 53 countries. Some have two. I chaired that network for the last three years of my time at the OBR, so I spent a lot of time talking to newly created bodies and ones that had been around for quite a long time. I was very happy to steal good ideas. We stole the approach, which I do not know whether they still use, to looking at the uncertainty around the costing of policy measures from the Australian Parliamentary Budget Office. Those engagements and discussions are very useful.

It is important, though, to bear in mind that these bodies differ quite a lot among themselves relative, for example, to central banks, which would look more similar if you compared them but different none the less. Of the 55, I think 18, including the OBR, prepare forecasts. The others do not. They would scrutinise the forecasts that were produced by the Government. Of the 55, 44 monitor fiscal rules. The others do not. Either that is not part of their job or there are not rules there.

Very interestingly, 42 out of the 55 make policy recommendations. They tell the Government what they think they ought to do, up to a point. The OBR, as somebody described way back in 2010, is very much at the positive and not at the normative end of the spectrum. It is quite interesting when you hear about the enormous power or influence that the OBR is exerting over policy decisions. It is very much in the small minority that do not make policy recommendations and do not give policy advice.

The one thing they all have in common is that they do not take policy decisions. There is one that is empowered to stall the budget process and make the finance ministry go back.

Lord Razzall: Which is that?

Sir Robert Chote: I think it may be Hungary, which is not in other respects a poster child example for a fiscal council. They differ a lot, and that depends on what you think the disease is that you are trying to cure. Where does power reside in budget-setting? Is it more with the Executive or with the legislature? If it is the Executive, is it the finance ministry? Is it operating in a system prone to coalition Governments?

There is then the question of: are there already existing unofficial bodies in these countries? One of the factors shaping the way in which the OBR was created was that we already had the IFS. There was no point simply creating a souped up, nationalised IFS. In some countries, obviously, supranational fiscal rules and fiscal scrutiny—particularly within the EU—are important. Here, even when we were in the EU, falling foul was probably regarded in some cases as more a badge of honour than a source of distress or concern.

You learn an awful lot, but these bodies differ in different places. It is one of the reasons why it is a very interesting group of people to talk to and learn from, but there is no one-size-fits-all model for us.

Lord Razzall: Do you think we are an outlier with regard to those other countries, or are we somewhere in the middle with what we do?

Sir Robert Chote: The most unusual feature is that the OBR is the monopoly producer of the official forecast, and the Government have to decide whether to take the numbers, how to respond to it and whether to publicly disagree with it, whereas in most other countries, it is the fiscal watchdog that basically is commenting on or, arguably, auditing the policies or the forecasts that the Government produce. That generates a whole different set of problems and issues. We get a lot of added value from the fact that the OBR does the forecast and does not simply opine on it from the sidelines.

If you go down that route, you are into big rows about what constitutes a reasonable or an unreasonable forecast. You get into the problem that many other fiscal councils have of securing access to information from government. The advantage of the OBR system is that the OBR has to be given the information, and it is a problem for the Treasury and for everybody else if it is not, because that affects the way it makes the forecast. Whereas if you were, in the OBR, not in the driving seat, you would be pleading for and trying to get information out of the Treasury, which would mysteriously turn up at the wrong time in the wrong format, unhelpfully, and so on. The framework and the institution have been under challenge—you do not need me to tell you that—over recent weeks. Moving away to that model will bring you a whole different set of challenges for the future.

Lord Razzall: Andy, do you want to add to that?

Andy King: One thing the OECD network is very useful for is when you face a new challenge. That network worked very well as fiscal councils tried to take into account the fiscal implications of climate change and climate policies over the longer term. The OBR went first, in some sense, in trying to do that, partly because the Climate Change Committee here produces such a lot of information that it was able to. The Irish Fiscal Advisory Council followed, and then a group got together to try to figure out how they could make that job easier for fiscal councils around the world. As it happens, the piece of technology and the model that they can all use has been launched literally today by the OECD, showing how working together can help everyone.

Lord Petitgas: Picking up on the discussion we had about the gilt market earlier, if the OBR were, in effect, to stop owning the central forecast, the Treasury took it over and the OBR had a new model in which it owned dynamic scoring and gave advice and commentary, do you think that would be a positive or a negative for the markets? Again, you will say, “If I knew this” but—

Sir Robert Chote: I would say it would be negative for transparency and understanding of the fiscal position more generally for voters, electors, Parliament and investors together. I think that would be true across the board.

The Chair: We already have the IFS doing that sort of function anyway, right?

Sir Robert Chote: Exactly. Obviously, I was running the IFS before I moved to run the OBR and it was very interesting because I had 10 years of trying to second-guess the Treasury’s forecasts—you may remember some of that period and it was not always terribly well received, as I recall, by the Government at the time. There are some bits that you can have a pretty good stab at. If you have good data and a good forecast for what will happen to wage growth or consumer spending, you can chuck the dart at the board with reasonable confidence around income tax and VAT. However, there are whole swathes of the forecast for which you cannot realistically do as an institution looking in from the outside.

In the model from outside, you obviously need an economic forecast to drive the fiscal forecast, because you need to know the things that are driving public expenditure and spending. I remember writing about how an OBR could potentially work, and having spent, earlier, 10 years as a journalist mocking other people’s attempts to produce economic forecasts, you think, “Surely the thing to do here is to basically piggyback on the consensus forecast or on the Bank of England to do that”. That does not really work because their forecasts are not fit for that purpose.

As we have seen, interestingly, in the factors that offset the change in productivity judgments underlying the recent Budget, the size and composition of nominal GDP are very important in driving your view of tax receipts. The Bank of England and outsiders are not producing a really well-thought-out, five-year-ahead forecast that focuses on nominal GDP to the extent that the OBR needs. That is another reason why doing it as an auditor or as a commenter would not be straightforward, because that is not a straightforward part of the process. You could not just say, “Well, it is unreasonable because the number is a bit low relative to the average forecast here”.

You will recall that in the Brown period, the concern over the over-optimism of deficit forecasts was around where you thought the proportion of GDP for financial sector corporation tax profits was going to end. There are not 17 other people’s estimates of the proportion of GDP that you will get from financial sector corporation tax to compare it to.

Q48 Lord Turnbull: Two families of models have appeared. Let us call one the participatory model, which the OBR clearly is, where the forecast is almost a joint venture, and the other is the scrutiny model. Your Which? summary was more negative on the scrutiny model than I thought it might be. You think the flow of information is the one thing that is probably better under the participatory model, but it has its disadvantages. You never know whose forecast it is. Whose forecast is it that real public expenditure will grow by only 1% a year? No one really believes it, but it is the policy. Are there any changes you could make that would take you closer to the scrutiny model but still maintain the benefits of the participatory model?

Sir Robert Chote: I would be very wary of blurring the two. As you say, it needs to be very clear whose the forecast is. It is very clear whose the forecast is, and the inputs that you put in from public expenditure in year 4 or 5 are policy variables that the Government provide to the OBR. As Andy said, those are sometimes not necessarily the most plausible assumptions, so you draw charts in the EFO that point that out. However, at the end of the day, you remember that the OBR is tasked with producing not an unconditional forecast but a forecast conditioned on what the Government’s policy is, taking into account those levers that they have firmly under control, even though the OBR knows and occasionally points out that, if you look at the past, the lever has tended to be pulled or pushed in one direction more than in the other.

Yes, I favour the participatory model, in which it is firmly clear that the OBR has ownership. Some people have said, “Why don’t you take the Government’s view of the economy and take the fiscal forecast off that?” Again, I think that would blur the responsibility. Keep it simple, keep it clear who is doing what and do not blur those two elements.

We have had variants of the scrutiny/audit model. Again, returning to yesteryear, Gordon Brown got the National Audit Office to audit particular assumptions that went into the forecast. The National Audit Office did not want that job, very sensibly. Ministers routinely overstated the imprimatur that have been given as a result of that, and it was a blurred and unclear model. As I say, this has not been an easy period for the current set-up and the current assigning of responsibilities, but you could achieve a lot by rediscovering the virtues of a taciturn period in the run-up to Budgets that would leave the strengths of the model well in place.

Baroness Wolf of Dulwich: You have been very clear about the advantages of being the monopolist—that the Treasury has to give you the data—and about the importance of it being very clear whose forecast it is. In the context of looking at other fiscal councils—not all 55 of them, but maybe some of the best established—do you feel able to elaborate on how these tensions manifest in countries that do not do it our way?

Sir Robert Chote: That is another way of saying, “What were the things, when I was involved in that network, that people complained about most over a glass of wine?” There were two areas where it manifested itself most frequently. One was whether they were appropriately funded to do the job. If you are the scrutineer, the temptation is always to whittle that away. If you underfund the OBR, you make your own life more difficult as well as the OBR’s, and that is not to be underestimated as a powerful alignment of incentives.

The other was access to information. I think in two cases, Spain and Canada, the fiscal council has gone to court to enforce its right to information from the Government. That is not a situation you want to get into. This a slightly separate issue, but if the participatory model, to use Andrew’s phrase, is going to work it has to be based on a good, functional working relationship. By the time you are taking your Government or your fiscal council to court, you are well past that.

The two areas that are most challenging are, essentially, funding and access to information, as well as complexities about appointments—delaying the appointment of new people and that sort of thing.

Lord Razzall: Does Canada also score the election manifestos? Is that Canada?

Sir Robert Chote: The “top of the class” model for that is the Netherlands. The CPB in the Netherlands does a fantastic and hugely labour-intensive job of doing the manifestos. The Australian Parliamentary Budget Office does it, but it publishes the results after the election. I cannot remember whether Canada does it. I am not sure now.

Lord Burns: I understand the case for the participatory model. As you will remember, for many years I was in charge of the Treasury forecast. The people who worked on that, including Chancellor Lamont, used to have regular discussions as the Budget process went on about how the forecasts were shifting, whether there was any new information, and dealing with the concerns that Ministers might have that one was being too pessimistic about this, that or the other. In this participatory process, as far as I understand it no real exchange takes place between the OBR, which is doing the forecasting, and Treasury Ministers. I sense that there is quite a lot of frustration on behalf of Treasury Ministers that they do not have any real opportunity to have that kind of exchange during some of the difficult moments. Is that a problem? Is there a way of dealing with that issue?

Sir Robert Chote: There is very extensive interaction between Treasury officials and the OBR through this, and I presume that Treasury officials are guided somewhat in what their Chancellors and Ministers are telling them. For example, the Treasury is present at the discussion on, “We’re going to sit down with HMRC and talk about where the income tax forecast is going and where the capital taxes forecasts are going”. Treasury people are there and are perfectly entitled, and able, to pitch in and say, “Wait, haven’t you thought this?”, and, “Surely that’s a bit of a dodgy judgment; what do you think about that?” The OBR takes all those suggestions on their merits. Similarly, obviously it is the Treasury that brings to the OBR a detailed assessment of the policy package, which may focus primarily on the direct costings, but there are also the meetings around the indirect potential effects on growth.

In my day, we had a couple of meetings with the Chancellor and with officials during the course of the Budget process or the preparation process that, depending on the tastes of the Chancellor in question, would focus on substantive judgments versus, “How are you going to explain this?”, or, “What are you going to say about that?”, which helps them frame the package. I think there are now more of those than there were in my day.

Andy King: Yes, absolutely. Over time, the number of actual formal meetings between the OBR committee and the Chancellor has increased. In my experience, they became more substantive over time. There were more conversations about forecast judgments or supply side growth judgments. The thing I remember—and I was in that team, not when you were running it but in between then and the OBR—is that some of the volatility between forecast rounds that seems to cause consternation among Ministers and their advisers is a function not of the OBR but of a multi-round forecast in a £3 trillion economy.

The fact that institutional memory does not extend that far is at risk of blaming the OBR for something that is fundamental to the forecasting process. If it were taken back inside the Treasury, that volatility would still be there. The difference would be that it could be managed by massaging forecast judgments. Having read a couple of memoirs and the rest, my impression is that people felt that massaging was not so much the Chancellor saying, “Revise up gross operating surplus, I need another £1 billion”, it was internalising the needs of the Ministers, because the officials’ role is to serve the Minister. At the margin, making those slightly more optimistic assumptions is more likely in the Treasury than it is in the OBR system.

There is a counterpoint to that, which has happened a couple of times in the past and it happened very much in this Budget. I do not think there is any way in the world the Treasury could have published the forecast that was just published, which said, “We have revised down productivity growth to bring our real GDP growth forecast into the range of consensus, but we have discovered that nominal GDP and its composition not only off-sets that but more than off-sets and we’ve revised up the tax forecast”. The fact that you have an independent forecaster and that is its best judgment allows that to happen. The Treasury just could not do that. Off-setting that, I do not think the Treasury could ever have got into the position that the OBR was in at the start of this year, when its real GDP forecast was the most optimistic medium-term forecast out there—swings and roundabouts, as ever.

The Chair: Colleagues, we will have make slightly more rapid progress.

Q49 Baroness Liddell of Coatdyke: I will build on the discussion that there has already been. I was never as senior as Lord Lamont, but I was a Minister in the Treasury, and when we got information it had come through so many sifts before it got to you. That was quite a difficult thing to deal with. Looking at the OBR today—well, in the past week—people who had never heard of the OBR now know about it. Is there anything that could be changed in the OBR’s mandate that would take it forward and take it out of the eye of the storm?

Sir Robert Chote: The key point is to emphasise what the mandate fundamentally is and to conduct the process more sensibly. It is a reminder that the OBR is there to produce the best forecast it can and to discuss the uncertainties around it, based on the policies of the Government of the day, and to provide the analytical information that is required in the run-up of fiscal events to allow the Government to make sensible decisions and for that to work well. What you need is a good, effective working relationship. You can write all the memoranda of understanding and legislation you like, but there has to be a fundamental trust between the two institutions. There are different interests within the institutions: there is an official Treasury, a political Treasury, No. 10, the OBR, and other government departments that may have differences of view with the Treasury on particular policy.

It is a complicated ecosystem, but a good working relationship has to be based on your ability to share information in a timely, candid and responsible way. From the OBR’s perspective, you want to do that in a way that is as far as possible consistent, not capricious, and does not try to make people’s lives difficult just for the sake of it or for entertainment. On the other hand, bluntly, you want Ministers and the wider political and official side to spend the weeks running up to a Budget concentrating on the substance of it and how you will explain it to people after the event, rather than trying to manage expectations as you go through or to provide a running commentary on either the policy package or the forecast.

If you did that you would get a long way back to the system working as it should do, rather than saying, “This has been a very difficult period, why don’t we chuck the baby out with the bathwater?” That would throw away what has been a remarkably imaginative—and set up by the Treasury—rather unusual but quite successful and fundamentally sound strategy.

Andy King: To add to that, I have been reflecting a bit on the Budget process itself that the OBR mandate sits within. There are parts of it that feel like they are just very difficult to manage today, when rolling news is a thing and so many journalists are calling the Treasury to ask leading questions so that even a “no comment” allows them a story, which I heard on a podcast somewhere. There are elements of that process that are unusual in the UK relative to other countries.

The fact that the Budget is an event—it is a speech that you wait for when everything is revealed—is very unusual. Budget processes really are processes almost everywhere else, where there are transparent stages. As that managing of the message has become more and more active, while still trying to have a Budget event that is an event, we have got to a stage where Robert’s model is to go back to a stage where the process is not transparent and the Budget is the moment of transparency. We are in a halfway house where there is a lot of uncertain transparency.

There is another model, where you bring a lot more transparency into each stage of the process. I wonder whether the events of the last few weeks allow a bit of broader reflection on whether there is a Budget process that allows us to keep the good of having an independent forecast. We were talking earlier about policing; the OBR polices analysis so that the analysis is unbiased. That is a really valuable thing that you want to keep, but it can either not talk about forecast judgments at all until the day or it could have revealed its productivity review—“At the start ask questions, reveal the conclusion of it”. There are pros and cons to all these things. I wonder whether they can all be thought through.

Baroness Liddell of Coatdyke: That is quite interesting. Thank you.

Q50 Lord Petitgas: Obviously this question is really from the past. How would you characterise the OBR’s relationship with the Government? You can discuss today or from your experience. Were there tensions in your day? How were they managed? What would you have changed about the manner of interaction, formally in the MoU or at the informal level?

Sir Robert Chote: I had several Chancellors in the period I was there, but generally speaking the relationship was a good working relationship. Both sides understood the role of the other and the fact that the other side was going to say or do things that would occasionally be inconvenient to you. There was a recognition that the process, in the end, served the public good well.

By and large during my period, you did not get briefing against the organisation or people talking about bits of the forecast in advance. There was a straightforward relationship built on candour. There were always bumps in the road: deadlines not adhered to quite as you would wish, or information coming late. Problems would arise; for example, if the Treasury was involved in a row with DWP then that would delay the point at which you were told things because they did not want information to flow. There were all those sorts of difficulties, and it comes back to the makings of a good working relationship, which is understanding the rules of the game, communicating in a candid, good faith fashion, and recognising, certainly from the OBR side, that Ministers and politicians live occasionally complicated lives under pressures that we do not confront; you live with that.

By and large, I think this is underpinned by the knowledge in the Treasury and the knowledge of most Chancellors that the system working well is in their benefit. Even if you do not always get the answer you would most like on a Monday or a Tuesday, looking back on it six months later it is the right way to be. Bluntly, it was easier for George Osborne because he created the model of the institution together, so the bar from having changed your mind and saying that you are clowns or frauds is probably a bit higher, but in general I think people have dealt pretty well.

As Andy said, there are all sorts of pressures on the news cycle and wanting to see more information, combined perhaps sometimes with a genuine desire to bring more transparency within the process. Having private space for discussion, for the sharing of a forecast and details of what the policy is, but then being as fully transparent as you can be when you come out of that about the judgments that you have reached, is the right way to go. Concentrate on the quality of the sausage you are presenting to the customer; do not get overwrought in wanting everybody to see every part of the process of making it.

Andy King: Speaking from my chief of staff of time, when I was managing the staff interactions with the Treasury, of which there are an enormous number because of the enormous number of forecast models and issues going on, I used to think that the clarity of the OBR’s mandate made that quite easy to do. The legislated requirements to be objective, impartial and transparent meant that any member of OBR staff going into a meeting with Treasury officials knows basically that their job is to try to find the right answer and then to explain it clearly when we get to document time. That meant, for the Treasury side, that they knew where the OBR is coming from. It has quite a simple set of objectives and that allows the process to work quite well.

I used to think that 99% or more of the time it worked well. In those moments when it did not, as chief of staff I would fire off an email saying, “Remember Clause 9 of the Act that says the OBR is allowed this access”. That usually meant we were dealing with a part of government that was not familiar with the OBR, so felt a bit nervous about sending off this policy under development, and it would all come together. The only thing I used to remind everyone was that the Treasury can deal with bad news much better than it can deal with surprises. As soon as we knew something, as chief of staff I would be on the phone to someone to say, “There’s something coming. We haven’t quite nailed it down yet, but it will be difficult”, so that when it arrived it was not a horrible surprise.

The Chair: On that theme of avoiding surprises, is there a case for things like the productivity revision, which is obviously a major change in forecasting, to be done much earlier than just in the run-up to a Budget?

Andy King: Yes, and I think that is exactly what the OBR did.

The Chair: Did it not do that in August, rather than in March? Maybe I am wrong; maybe it did not.

Andy King: It did it in August, which, to my mind, is very early for a November Budget. It finished the March forecast, did a couple of other things and then moved straight into that supply side review to get it done as quickly as possible.

Sir Robert Chote: Crucially, it also pointed out in July, in the Fiscal Risks and Sustainability report, that the productivity forecast was relatively high compared with those of the Bank of England and other forecasters. I do not think it took a work of genius.

The Chair: It was signalling.

Q51 Lord Liddle: My question is about whether the way the OBR communicates with the public could be improved. One of the things I feel is that the OBR does a lot of very good work on the long-term issues to do with public spending but these very rarely seem to get the attention that the short-term stuff about Budgets does. Could that be improved in any way? Then, is there any merit in trying to be more open about the process of successive iterations on the Budget? I think we were told that there were five or so iterations of the forecast on this occasion.

Sir Robert Chote: On the former, the OBR has done a lot—a lot of it since my time—in trying new channels and ways of communicating, through videos, social media activity et cetera, about some of those key things, particularly out of the Fiscal Risks and Sustainability report. One challenge with the fiscal sustainability report is that the conclusion is generally the same: that, on current policy reasonably defined, you are on an unsustainable path because of health costs, the loss of some tax revenues through loss of fuel duty, et cetera. You are not saying new things about that core element but there is a lot of important stuff. The work that Andy and colleagues did on net zero has had a lot of purchase.

Quite a lot of it, it must also be said, is quite complicated, and some of it is not the most exciting in the world to the person down the Dog and Duck. You also have to remember, in what the OBR writes, produces and communicates, that you are dealing with different audiences who are all important but have very different focuses, so you need to meet the needs of the social media journalist who has to summarise the whole thing.

Lord Liddle: I am talking about the political class when I ask this question.

Sir Robert Chote: We had discussions at various points about, “Goodness me, aren’t these reports just getting too long?”, for example, and whether people were getting bogged down in that. We tried to focus so that you have a snappy overview, a more substantive but still digestible executive summary and then more detail. I always thought, though, on the length of detail, that the Treasury always needed to know that there was no part of the public finances too boring or complicated that, if it tried to get up to dodgy tricks, the OBR would not produce a two-page box explaining what was going on. There is that value of doing stuff that is not simply aimed at the most casual observer.

Remind me: what was your second point?

Lord Liddle: The second point was about the iterations. Is the conclusion of the latest episode that this should be more transparent?

Sir Robert Chote: Personally, I think not. There is a virtue in having that space for private interaction and iteration of the forecast. That is partly because producing the forecast and the policy package and, for the Government, the communications around that is complicated enough without having to worry about how we will explain why it moved from round 1 to round 2 in the way that it did and whether there anything suspicious about that, or whether the Government feel they need to have an explanation for why they were thinking of policy A and ended up with policy B. I would not go into that, personally. I can understand why they felt that it was necessary to publish it in that way but I would maintain it in that fashion.

Another thing to bear in mind is that when you look at the estimates of the bottom line of the forecast in round 1, round 2 or round 3, it is easy to assume that the reason why they differ is because the OBR has had a change of heart about something between one round and the other, whereas an important part of the iterative forecast process is that you start making big judgments and colouring in the broad brushstrokes. Then, as you get through to later parts of the forecast, you fill in more of the detail. You will not have a particularly well-articulated view of the housing market forecast or of the forecast for financial sector corporation tax receipts in the first round.

Part of the reason why the forecast evolves from round to round is not that you are changing your mind but that you are producing a richer picture that is based on more inputs, more consideration of what you have heard from the different bits of HMRC, DWP and so on. Therefore, putting too much weight on why it went from £2 billion in round 1 to £4 billion in round 2 to £3 billion in round 3 just adds greater complication to what is an already complicated framework. One lesson of recent events is to keep it simple.

Lord Liddle: Would producing a range of outcomes or emphasising more the range of probabilities in any outcome help the usefulness of what the OBR does or just mean that people give up and think it does not really mean anything very much?

Sir Robert Chote: We had a chapter in every EFO under my watch, and I think in every one subsequently, talking about different ways of explaining and understanding the uncertainty that lies around the central forecast. You can look at past errors or past differences between outturn and the forecast; you can look at sensitivity to particular judgments; you can look at the uncertainty in various ways and place probabilities or talk qualitatively about that. I tend to be pretty wary of ranges, because I think they just provide you with two numbers that people can focus on and decide which is the more newsworthy, rather than one in the middle. I am inclined to be very much in favour of talking about and explaining uncertainty, but by saying that the chances are that it will lie between X and Y, everybody will focus on whether X or Y serves their interest better or makes a better story.

Andy King: We tried so many ways to take away the emphasis on the single number headroom, but the fact of the matter is that that is what people care about, and if it is in the book it is the thing that will be focused on. The number of times I read that the OBR should do probabilities or fan charts or something that is there already: “It should do scenarios”: there are loads of them; “It should think about sensitivities”; there are lots of them. It is just that that is what people care about.

The one example of an institution that tried very hard to stop people focusing on a single number was the Bank of England in the Mervyn King era, where it was not in the book, so the only way to find out the inflation forecast was to download the PDF, blow up the picture of the inflation forecast as big as possible and dot it on, which is exactly what all the City analysts did: they created the inflation forecast off a picture because they needed a number. In the end, the OBR provides all the analysis of uncertainty—that is the right thing to do, and the probabilities and so on—but it also transparently sets out how that headroom number moves, because that is what people care about.

Q52 The Chair: As a quick follow-up, you alluded before, Robert, to the frustration that the EFO gets all the attention and the Fiscal Risks and Sustainability report gets less. Can anything be done to raise the profile of the Fiscal Risks and Sustainability report as a major economic knowledge moment, if not a fiscal moment, from a government point of view?

Sir Robert Chote: The intention was originally to make that a robust element so that the Government had to respond formally to those reports. I think it is still the case that they have to do that. If you look back over my time, there were periods when there were detailed, enthusiastic engagements with that report and there were some that were rather more terse and rudimentary. The extent to which there is willingness to engage on that from the Government’s point of view is clearly true. Then, of course, there is the value of the committee in the other place, maybe—it regularly deals with the dais of the reports, rather than on specific enquiries—taking those seriously, asking questions about them and asking other people for views on them. That would be helpful.

Lord Blackwell: The problem with people not taking the probability distributions is because it is the OBR that is responsible for producing the number for the forecast, in the same way that the Bank of England has to produce the inflation number that people feed into their models. If the number came from the Treasury, the OBR could avoid having a number and could simply say, “Our view is there is a range of probabilities around it”.

Sir Robert Chote: You are suggesting that the Treasury produces the forecast and the OBR tries to put a—

Lord Blackwell: Yes. If we were in that alternative model of the Treasury producing the forecast and the OBR auditing it, the OBR would not have to come up with a number.

Sir Robert Chote: It would depend on how robustly you were trying to answer the question. If you were trying to answer the question robustly, you would say, “Here is my central forecast and probability distribution. Let me lay the Treasury’s one on top of that and see where it falls, and that is telling you the answer to that question”, in which case you are back with the fact that you are responsible for producing a robust, transparent, well-explained central forecast and probability distribution. I do not think that would get you out of it.

This is another variant on the discussion on the participatory versus the scrutiny model. It is, and always has been, within the rights of any Chancellor to say, “I’m going to produce or publish an alternative forecast”—or, short of that, a detailed, well-argued explanation—“of why I think the outcome will be different”. I can imagine some past Chancellors with whom I dealt who would have had no compunction at all in saying, “The kids in the office have come up with this; I think that the outcome will look rather different from that”. The OBR is not forcing people in terms of policy, or indeed to accept that forecast. If somebody says, “I think my policies will be transformative on productivity to a degree that the OBR can barely dream of”, then by all means write a paper and cite some evidence for why you think that might be case.

Q53 Lord Davies of Brixton: I am eating into my own time but, Andy, you said that the model is extremely complicated. I am very suspicious of complicated models because you end up not having the faintest idea what it is actually doing. Have you done a thorough think through as to how complicated the model needs to be?

Andy King: Absolutely. There is no “the model”; there are many, many models. I am probably a bit out of date now, but I think the OBR’s economy model has roughly 500 variables in it. It has so many because they are individual determinants of taxes. If you want to forecast tax on savings income, you need to forecast savings income. No one else out there doing economic forecasts bothers with that. All those things exist in the economy model. In a way, it is not complicated. It is large.

Lord Davies of Brixton: That is what I was wondering.

Andy King: It is not sophisticated in the sense of central bank forecast models. It is large. A lot of what it does is ensure that A is consistent with B because B is always half of A, so make sure that happens.

There are then lots of individual models for individual lines of tax and spending. When we totted them up in my day, it was a little over 300 different models. Some of those are incredibly simple spreadsheet models and some are incredibly complicated. The income tax model is a microsimulation of a representative sample of taxpayers run by HMRC that the OBR is not allowed to look at because it has confidential data in it. It is necessary because any time you want to change the tax system, you will change it for a certain type of person with a certain type of income.

All these things are horses for courses models. They are designed to be as complicated as is necessary and as simple as is possible. It is a constant work programme of improvement, of course, but it is designed how it is because it needs to be. It also comes back to whether you want to constantly give a running commentary on those pre-measures forecast rounds where X feeds into Y and needs to feed back into X. You do not want to explain all those, “Only update models 57 through 92”, in round 2 because you are busy doing the first ones in round 1. It is complicated, but because it has to be.

Sir Robert Chote: There is also a slight danger that the model can sometimes have a mythic quality to it. People have this vision of some vast neoliberal supercomputer humming away in the basement of the Treasury that is driving ideologically driven conclusions out of it. As Andy says, a lot of it is making sure that things that add up do add up, ensuring internal consistency, reflecting relationships between stocks and flows, and ensuring that you are taking into account some plausible things you think you know about behaviour.

However, most of the big judgments on how much the economy will grow over the next five years do not fall out of the model or some mysterious function of it. That is the BRC sitting together, looking at the evidence, how much spare capacity there is now, how much we think the underlying potential of the economy will grow, and where we think the actual activity will be. Some of that big picture stuff is based on judgment and the model, as much as anything else, ensures that the story that you are telling that you want it to be consistent with is internally consistent in its own right and explicable.

Andy King: The most important assumption—the productivity one that was conveyed in August and that 60-page paper looking at it—is a decision that is taken outside the model. That decision is then plugged into the model to determine GDP growth, as an example.

Q54 Lord Turnbull: Robert, you set out some benefits that you thought would be lost by moving to a pure scrutiny model, in particular the sharing of information. What can you do while retaining those benefits and do they add up to a sufficient improvement?

The first has already been announced: we will only have one Budget event a year. We will work with larger headroom, possibly even larger than now, with less prominence on hitting a target in a particular year. All those three will produce less twitchiness in the policy process. The other would be avoiding, to use your phrase, a running commentary and kite flying, and having greater discipline in both parties, particularly on the Treasury side. The fifth one, which the Chair has referred to, is that the Fiscal Risks and Sustainability report should be enhanced or restored to its status and possibly used as the occasion to unveil a longer-term study of productivity, ageing, climate or whatever. The question is: if you did all those things, would it add up to a sufficient response to the difficulties we have encountered this year?

Sir Robert Chote: It would add up to a decent response to that. As I said, the fundamentals around the value of the transparency and the vigorous tyre kicking and scrutiny that goes in at all the meetings that Andy has described at official level and with the members of the BRC involved are a hidden benefit of the process as it works now, and it is necessarily hidden. It is in the machinery itself, but it is an important one.

There had been some speculation, for example, about not merely going down to judging the fiscal rules only once a year, which I will come to in a minute, but having only one forecast a year. That would have been a bad decision. It would have been a step back in fiscal transparency. It would have meant that you were turning the handle on it only once a year, with no chances of getting bigger changes from year to year than you get now, when the bonnet has not been lifted up on the HMRC models that Andy has described for the last nine months and then you suddenly describe you see the tangled wires. I am glad that they have not gone down that route.

It is worth saying that that has been framed as saying that the OBR will now be required to assess compliance with the fiscal rules only in the autumn rather than in the spring. That is an odd way of framing that, because the OBR will still produce a forecast of the variables that the Government are targeting, and any idiot will be able to open the book and say whether the OBR says that you are on course to hit the forecast or not. That announcement means that the Chancellor is saying, “I will not necessarily take action in the spring if the forecast shows me missing my rules”. That is, as you say, a way to try to get away from the fine-tuning of fiscal policy in response to forecast movements.

However, it takes you back to an old dilemma about what happens when you publish forecasts that do not show the Government on course to achieve it. I often refer back to Alec Cairncross, head of the GES. When he stepped down, he gave a lecture to the Royal Economic Society in 1969 and asked why Chancellors have historically been anxious or nervous about publishing forecasts even when they are responsible for them. He said, “Error is not the main hostage that a Chancellor thinks he is giving to his political opponents. It is rather that an official forecast becomes in the act of publication a plan”. Will you be able to say, “There is a forecast, but that is not how I want things to turn out”? You then have six months of saying, “Okay, Minister, the forecast shows that you are not on course to hit your rules. What are you going to do about this?”

It is a sensible step to try to get away from having the policy being as variable as the forecast, but the idea that this will mean that there is no speculation for six months a year about what will be in the next Budget is for the birds. It is important that information is there, and you would not even need the whole forecast to be there, but you are not going to get away from a world in which the Chancellor or the Prime Minister is asked at various speeches or events and so on about future tax developments. We need a sense of realism about how much that particular announcement will contribute.

Andy King: The crucial thing was increasing headroom: it is not the downplaying of the spring forecast, it is having enough headroom to cope with a revision of the spring forecast. If we assume that the OBR is unlikely to revisit major supply side judgments between the autumn and spring as well, so one source of large forecast revision is taken off the agenda, you could see that headroom plus a slightly different calendar gives you enough to mean that that scenario of being off course immediately is less likely. I noted that the Treasury document said, “We won’t respond unless we need to”, which is quite a pragmatic way of trying to downplay the spring forecast.

Q55 Lord Davies of Brixton: You have answered my question. Some have called for the OBR to produce only one forecast per year. How many do you believe is optimal? You have said two.

Sir Robert Chote: One would be a really bad idea and two is sensible.

Lord Davies of Brixton: You described the model, or the system; we will call it a system rather than a model. You have this large bit with the 500 variables, which is a fairly deterministic set-up. These figures come in from other sources, you put them in and you get there. Alongside that are some other decisions. Presumably, the board of the OBR takes a view on these different things and then they are melded together.

You said that some of the decisions are made only once a year and others are made twice a year. As a thought experiment, could you produce a forecast six times a year, say, each time? You could then see more readily where the economy is going. I am not suggesting that as a serious proposal but as a thought experiment: would it in some ways work better because you would see the bend in the road before you hit it?

Andy King: Diving into the weeds of the models might have given the wrong impression of what happens. The forecast is entirely the construct of the committee. The committee makes all the decisions. There is one big model and then lots of other models. They are the tools that are used by the staff to turn the committee’s judgments into a forecast. They are also the tools that are used by the staff and officials to inform the committee to make those judgments. They are very much tools. They do not produce the forecast; they are accounting tools, econometric tools or whatever.

With sufficient resource, you could be forecasting constantly. For those of us who were there in the summer of 2022, when the Conservative Party leadership debates revealed that a call for an emergency Budget was very likely, the OBR started forecasting in July and kept going until, finally, there was a fiscal event with an OBR forecast in November. There were seven or eight rounds of forecasts. Conceivably, they could all have been published; they were at varying degrees of completeness. You could have a constant running commentary on the forecast. There is a monthly running commentary from the OBR on how the public finance data in the year are performing against its in-year forecast. That would give you some of it.

The problem is you could have too much of a good thing. Yes, you could see the turn in the road coming, but the Treasury can do that itself. It can monitor the data internally, producing a lighter version. It will typically focus on the economy forecast and a simple ready reckoning of the fiscal consequences. It can and does do that internally. Having that published with the OBR kitemark on it every month or two might get into diminishing returns or creating new problems through excess information.

Sir Robert Chote: Also, it is worth emphasising that the OBR does not know anything about the evolution of the economy that other people do not know; it has the same access to the same information to read judgments on that. What is different is twice a year interrogating the people who are trying to do the North Sea oil forecasts, for example, the different bits of the corporation tax forecast, or local government finance—which was my favourite, as Andy knows, over many years.

You get flows of information at different points of the year. There is a point at which you get self-assessment data coming in. At points in the year you get information from local authorities. There is not a constant set of complete data. The real world is evolving all the time, but at points in the year you suddenly know more than you did beforehand about how departmental spending is performing relative to what the Treasury had been assuming, what local governments are saying about the size of their reserves, decisions on self-assessment income tax, or how many estates are coming up for inheritance tax, et cetera.

An awful lot of the value of what the OBR does is taking the macro forecast, looking at the individual elements of the fiscal forecast and probing that. You want to get the DMO or HMRC to have time to go away and think about how they are doing it and say, “Have you looked at this? We’re not sure about this assumption”. They may come and say, “Actually, we’re not happy with how this model works. We’re thinking of doing it differently. What do you think?” An awful lot of that work takes place outside the heat of the last six weeks of a Budget round.

Lord Davies of Brixton: The market consensus does not work because it is partial; it has different objectives. You could not rely just on what the market thinks will happen.

Sir Robert Chote: No, it is a combination of the fact that people are primarily focusing on real rather than nominal GDP. Many do not forecast over a full five-year period. Even the Bank of England is still going out over only three years. I remember early in our day, when we were wrestling with the problem of there appearing to be quite a lot of spare capacity in the economy: how quickly do you think the economy can grow? What is your assumption about potential GDP? How do you get all those things to work together and cohere? That is quite hard. You cannot get that from looking at the average of outside forecasts. They do not go far enough; they are not triangulated like that. The Bank did not have to explain how all those things meshed together and we had to surface those sorts of judgments because of the nature of the forecast we were doing.

I was quite keen, before the OBR started, to say, “Macro forecasting is a mug’s game. Let’s try to get the blame for that on to somebody else and focus entirely on how the fiscal forecast evolves, but no macro forecast does that for you”. If you want to say intelligent, well-founded things about uncertainty and what underlying assumptions are key to the outcomes of the forecast that you are getting, you need to have a forecast that is fit for your purpose and be able to use it yourself.

Q56 Lord Turnbull: I have a quick factual question. Where does the OBR get its staff from? It started as a cutting from the Treasury, but where do people come from now and where do they go to? Is there a community of economists who are not Treasury people who you have borrowed? How far do you have an independent group?

Sir Robert Chote: Andy will have a more recent view. When we started it was certainly taking a set of posts and people out of the Treasury’s macro and main fiscal forecasting teams and sticking the committee on top with some additional support function. Subsequently it has grown, partly as the responsibilities have grown. The Treasury is still an important source, but places like HMRC and the DWP are an important source, and from outside altogether, plus new graduate entrants.

Andy King: Yes, over time it became less people with a Treasury background and more people from all over government and outside government. It is very much a diverse pool that is drawn from. Even the senior levels now, which were for a long time those original people, are less so. It is a diverse group now.

Q57 The Chair: You have been incredibly generous with your time. Can I ask you one last question? We have talked a lot about, as Robert put it, the pressures in the relationship between the OBR and the Government in the last few months. The last week has had that and then some with the things that have happened, ranging from the unfortunate leak of the outlook to the letter from the OBR to the Treasury Select Committee with the timeline, the Treasury annoyance at that, and then the resignation yesterday. We also have Ciaran Martin’s report about the OBR’s management of the outlook technical issues. You are not sighted on all these issues, but do you have any reflections about any of those events and whether we need to learn lessons about that going forward?

Sir Robert Chote: Let me start. Clearly, the leak was a serious and unfortunate episode, and it is absolutely right that they went straight to getting somebody independent to come in and look at that, and to make recommendations for how to address it. It is not my area of expertise, but Ciaran Martin seems an excellent person to have asked those sorts of questions. I am sure the OBR will be working together with other parts of government promptly on doing that.

I should take this opportunity to pay tribute to Richard Hughes for the work that he has done over the last five years since he took over from me. He has done an awful lot of great value for the work of the OBR. His decision to step down based on what he thought was best for the future of the organisation is entirely in character with what I expect and consistent with the spirit of public service that he has brought.

Stepping back further from that, it is a time for everybody involved to reflect on how you make this an orderly process in which the OBR is able to serve the public and the policy process as best as possible. The sorts of things that we have discussed so far—ensuring that the attention is focused on getting the substance right, being able to explain that after the event to Parliament, the public and financial markets, and doing that in as orderly and constructive a fashion as possible—are much to be wished for. I am sure that is what everyone will be thinking about.

The Chair: Andy, did you have anything to add?

Andy King: I agree with all that. Richard’s commitment to the goals of the OBR was visible every day. As I said, as I reflect on the last few weeks, this feels like an opportunity to reflect on the Budget process more generally.

The Chair: Great. We have no further questions. Thank you both so much for your time. We are grateful. With that, the committee is finished.