Welcome to the Treasury Select Committee on Tuesday 19 May 2026. We are here continuing our conversation about the role of the OBR in light of the loss of the chair of the OBR at the end of last year. There is recruitment going on for the chair of that organisation, but it is 15 years on, so it is time it was looked at. We are today delighted to have two panels. Our first panel is looking at international comparisons. I am really delighted to have Laura van Geest, who is the chair of the executive board of the Dutch Authority for the Financial Markets. She is appearing virtually with us today. Also appearing virtually is Professor Doug Elmendorf, who is the professor of public policy at the Kennedy School of Government at Harvard University. Thank you very much for joining us. To start, you have both led independent fiscal institutions in your respective home countries. What are the main differences between the institutions that you have led and the Office for Budget Responsibility in the UK?

Thank you for having me. There are lots of comparisons between the CPB and the OBR, so comparable things. We both make forecasts for the economy and the fiscal position. The mandate of the CPB is broader, so we also do other things than just the work that the OBR does. In the Netherlands, the CPB makes the macroeconomic forecasts and also makes a forecast of the fiscal position. Then the system is that the Government have to use the macro forecasts but can make their own fiscal projection then. Then there is the State Council, which does the assessment of the position. In the UK, the OBR makes both the macro forecast and the fiscal forecast—and that basically is the forecast that the Government use—and the OBR does the assessment. In the Netherlands the assessment is put aside and done from a different angle. That is basically the difference in the work mandate. Also, in the Netherlands the CPB has more leeway in how to go about making the fiscal forecast. In the UK version, there are requirements around stated policy. You have to make the assessment of whether the target will be met. The CPB has more leeway on how to go about its forecast and its assessment. These are two major differences. The other thing is that the whole set-up is different because of course the fiscal policy and the political arrangements in the Netherlands are different from those in the UK, and there is an interaction between the OBR and the political arena. There is also a different way of interacting between the Dutch setting and the CPB. Those are the three main things to highlight.

It is an honour to be here with you and Laura. The biggest difference between the Congressional Budget Office and the Office for Budget Responsibility stems from the different political systems in which they are embedded. The CBO was set up to serve a Congress that is separate from the executive branch and because of that CBO does its own economic, revenue and spending forecasts. It works with both the majority and minority parties in the Congress. The whole idea was, and remains, to provide the Congress with a source of information that is independent of the executive branch. That is the most fundamental difference. Maybe because of that, or maybe because we are a larger country, or some other factor, the US Congressional Budget Office is also much larger than the OBR. It is more than five times the size in terms of staffing.

It is still an important body that is respected and listened to, even though it does not provide the official forecast. What do you put that down to?

I am not sure what we would call the official forecast in this country. The Office of Management and Budget, which reports to the President, produces an economic forecast. CBO produces a different economic forecast. Neither, in a sense, is the official forecast. The Administration say what they want and the CBO serves the Congress differently. It really depends on the credibility of the organisations. CBO has a now 50-year history of providing credible forecasts. They are not always right, to be sure, but they are credibly independent forecasts. Different Administrations come in and then build or lose credibility, and they get more or less attention in the press and in the public debate depending on the political circumstances of the time.

Most countries do not have independent financial institutions also producing the Government’s official forecasts. The Netherlands and the UK are two of the three that do that. What do you think are the pros and cons of having an independent body do the official government forecasts?

The advantage is that independent fiscal institutes have less of a tendency to come up with a biased forecast. They are less stressed to look at the bright side of life, so having an independent forecast is a major advantage there. In the Netherlands, the independent forecast on the macro side is done by the CPB and then the Government have to use it. That is the difference between the Netherlands and the OBR, because the OBR also makes the fiscal forecast that is used by the Government. In the Netherlands, the Government just use the macro forecast, so the growth, unemployment and interest rates forecast of the CPB, and then come up with their own forecast for the Budget. The CPB also does a forecast for the Budget. There are differences, although generally these are slight differences. That is something that people then can debate and evaluate. The main advantage is that you have unbiased forecasts and so you have a better basis to make policy, because people tend to be rosy, normally.

Sometimes our Office for Budget Responsibility has been attacked for being partial when making judgments. There has been a groundswell of criticism under different Governments at different times. Do you have that same problem in the Netherlands and do you think that that is just the politics?

As an independent fiscal institute you have to be conscious not only to be independent but also to seem impartial. This requires a good way to talk, so it is basically using language, presenting analysis and not making judgments. In the Netherlands we have spending ceilings. If you go over that, so the ceiling is 200 and you are going for 202, that is where the CPB stops. Then the State Council says, “You are going over and that is not appropriate”, or it can say, “We understand”, so judgment is separate. It is a thin line and it is basically up to the chair, of the BRC in this case, or of the CPB, to be prudent in how to present things, to be analytical in their presentation and to make sure that people do not see them as partial. I did the third external review of the OBR. When we asked people how they experienced the OBR, in private at least they said that they thought it was independent and impartial, but there is a tendency in the press, because the OBR is the producer of the official forecast, when you have a problem with fiscal headroom, to say, “You don’t get any room” or, “The OBR puts you in a bind. You can’t move”. That is not right, because the fiscal framework is designed and decided by the Government, and the OBR just presents the facts. Of course, if you do not like the facts it could be attractive to blame someone else. That is the problem. The fact that the OBR also produces the official forecast makes that risk greater than in the Netherlands, where we, at the CPB, make a forecast for the spending review, or the ceiling, and the Government make another, so it is less that you are the one who owns the figure. That is a difference where I think in the Netherlands I was in an easier place.

This is the final question from me. There has been some discussion here at the fringes about whether the Office for Budget Responsibility should be asked to score manifesto promises of opposition parties. It is a very small institution here, so changes might need to be made. I wondered whether I could get both of your views on whether you think that that is a good thing. It is not common. The Australians do it, but otherwise I do not think that anywhere else does it. What do you think about that?

In the Netherlands we have a long tradition of analysing election manifestos. That is something we do. There is an export product of the CPB and I really like it. It fits into the Dutch system where everyone is always in a coalition agreement, so you need to have your policies thought out before the elections, because after the elections you have to go and negotiate. There is a big enthusiasm to do this on those sides and the CPB is respected for that. In the Netherlands we only have the CPB. In the UK, you have the IFS, NIESR and other institutes. I am not sure whether, in the specific set-up, also because the OBR really leans on the Treasury, that is fitting in the UK setting. We also tested this question during the external review and people were really hesitant about this because they were afraid that it would make the OBR seem more partisan. For that reason we advised against it because we thought that it did not fit into the UK system. I think that the IFS already does some work on this. It would be useful if someone in the UK were to be willing to do costing of manifestos or to produce technical assistance in that area, but probably not the OBR.

I share Laura’s scepticism that the OBR should do this. In the United States context, CBO provides estimates for both majority parties and minority parties at that time in the Congress, so CBO works both on the policies that are likely to go through and on the policies that are not likely to go through. That is different from doing things that are campaign promises. Also, just as a technical matter, at the CBO it matters that the policy is written down, word by word, in legislative language. Campaign promises are rarely so clearly spelled out, so that gives you more challenge, at least in our context. It makes it more difficult to evaluate a policy in a purely objective way if you are trying to fill in some of the details that may not be captured by the headline point. It would put the OBR in an even more difficult political position if it was tasked with that sort of work as well.

Thank you, Professor Elmendorf. That was a very diplomatic way of describing political campaigning.

Like I said, in the Netherlands we have a long tradition of assessing election manifestos. In our case, it is quite detailed. If we assess a policy, they have to come up with a very detailed description of what the policy looks like, because otherwise you end up with the problem that Professor Elmendorf signalled, so it is also a time-intensive process. It takes months, so it does not really fit into your system. In the Netherlands, people say that it is like playing soccer against the German team. You can lose it if you do it not properly with the CPB, but you hardly ever win the game through this process. It is not required in the Netherlands. Parties that undertake the costing exercise find it far easier to come up with compromises when they have to negotiate a government coalition agreement than when there are not any election platforms assessed. In the present Government, most parties had participated and it was easier than in the former Government, where there was only one party that had its election platform assessed. One thing that helps political parties in the Netherlands, and it is something that the OBR does not undertake, because apparently it feels that it is not allowed to, is that we also have booklets that are called Promising policies. We have a set of policy ideas that have been ventilated during policy discussions or have come up in economic journals. We say, “If you think you want to stimulate labour supply, you can undertake this tax measure, that tax measure, or such a tax measure, and this will be, broadly, the effect”. The CPB is quite influential in that way. It translates economic literature into possible policy measures that political parties can then consider in designing their platform. The trick for the CPB is then to make sure that, if you design those policies, so these Promising policies leaflets, you cover every type of political party you have in the Netherlands, so everyone can find something to their liking. In that way, you produce more political debate on policy and fiscal policy measures. That then feeds into the programmes and into government policy. That is one of the ways it works in the Netherlands.

That is very interesting and quite different. I can already feel the clash between our system and that proposal as it stands.

Could we move now to the UK fiscal framework? Could you tell us what you think about the UK’s fiscal framework as compared to other countries? What are the ways that it could be changed to produce a less contentious circus of commentary every fiscal event?

You have fiscal rules, and if you sail close to the wind there will always be debate, whatever fiscal rule you design. That is basically a fact of life. You could consider whether you have the best type of rule there is, because, to an outsider, it looks as though all the discussion is focused on the headroom five years out, generally beyond the moment that the Government will still be in place. It is also a moving target, so the fifth year you have to go down with debt, but the fact that you go up over time is ignored, which does not make good policy. If you compare it to the Dutch system, one thing that is different in the Netherlands is that people extensively discuss the fiscal framework they want to use during the government period at the beginning of the Government. That is, of course, part of having a coalition, but this does not seem to be that much the case in the UK. Of course, if you do not discuss it very intensely, you do not feel the same level of ownership. That is one thing I would advise and it is always good anyway. In the Netherlands we have a big focus on the short-medium term, so the Cabinet period, so four years out, and then we are very much conscious of the fact that forecasting is difficult. We have an expenditure ceiling, where we find it easier to predict what we will spend our money on and it is also less sensitive to the state of the cycle, and we let the automatic stabilisers on the revenue side work. We have a different system. We do not focus on the deficit or the surplus, but we have an expenditure line, so kind of the DEL but in addition all other expenditures, and then the revenue: “Does it work?” It means that sometimes the deficit will be higher and sometimes lower than anticipated at the beginning of the governing period, and you accept that.

Could I try to build on this? This Committee is trying to develop sensible recommendations for our Government here in the UK and it is really helpful to have your description of your own system. When you look at ours, I am sure you are familiar with the IFS’s traffic light system proposal and the reaction of Richard Hughes to that. What enhancements or changes, if you are not going to change the fundamental fiscal framework, could you bring in to try to create a more considered view of the OBR’s judgments? Is there anything that we could suggest that would remove this binary decision that “the OBR has said X and the Government have failed”? What do you suggest would be a refinement that is reasonable, given the situation we are in politically at the moment?

If you want to keep the headroom and so on, which will make life difficult, we said in the third review that you could give the OBR the feeling of more room to manoeuvre in presenting not only the point estimate but also scenarios and the uncertainty around it, and to give an indication of long-term sustainability. That is what we said. If you look at the press release, it presents growth and the fiscal forecast. In addition or instead, you could present the fiscal forecast with the uncertainty mapping around it, scenarios that could be likely and change this, and the long-term sustainability. That is something that we advised. In terms of the IFS stop light proposal, I am sceptical, to be honest. Having a general “motherhood and apple pie” text will not help the Government steer their Budget in the proper direction. As a former Ministry of Finance official, I would not have thought that that would be that practical. On the idea of the lights, so using lights as opposed to figures, I also have my own bad experiences with those. Just to give a sense of how this works in practice, in the Netherlands, the last time that election manifestos were assessed, they did not go with decimals but they had arrows, so up or down. Then the parties of course realised that you go up if you have 0.1 or 0.2, and down if you have minus.

The politicians muck it up as ever.

No, I do not say that they muck it up. It looked lovely. In practice, you have to say at which point you move from red or from green to yellow. You cannot just flop around and say “expert judgment”, and then do this, with things that are basically figures. Then it will be manipulated. It is quite difficult to do this in a proper manner. The ideas of the IFS to highlight the uncertainty around certain figures and to have a view on sustainability in the long run are useful ways to look at things. Having stop lights or colours will be defeated in practice, in my assessment.

The United States had numerical fiscal targets some decades ago and that system broke down because of the difficulty of making forecasts and with unexpected events occurring. Moreover, it was a period in which our Congressional Budget Office was under particular scrutiny for its forecasts, so I am sympathetic with your situation and the situation of the OBR today. The best thing I can propose is to soften up the rule to allow for a buffer in the way the rule is described and used, or to build in more headroom. As the IFS said, it is not possible for a fiscal rule to be simultaneously simple, flexible and enforceable. I do not think that the stop light approach really works. Laura has that right, in that people learn how to game a system. Moreover, if you have a lot of indicators and you can always point to the one you like, you have sort of lost the thrust here. Of course, the OBR can and should report different sorts of measures, and people can look at them, but, to have a rule, simplicity really is important. As she said, any time you have any kind of rule, there is going to be some effort to be close to it without going over. That is unavoidable, but the best approach is to recognise that uncertainty to begin with. The OBR is a leader in how it talks about uncertainty and the attention it gives to presenting uncertainty. I would try to build on that in its projections and think about a rule that is not exactly £100, but is £100 plus or minus 5% or something that recognises the intrinsic unknowability of the future.

On that, do you recognise the way that the OBR is slammed for almost being an arbiter of government policy and having the right of veto? That is the way it is presented in some of the political discourse. Has that happened in your country? How do you think that should be handled?

Some years ago, a senator became well known for saying in a hearing to the then director of CBO, “You are not God”. Yes, point taken. Nobody at CBO thinks that they are. It comes back to the fact that the policymakers own the policymaking. CBO has been very clear from day one that we are not the policymakers. We should not be the policymakers. When people say, “Nobody elected you, Director Elmendorf”, they are right. That is why CBO never makes a policy recommendation. It just offers analysis of what alternative policies might produce. The responsibility rests with our elected leaders, and thus with you in the UK. People can complain about CBO’s power, but that is not right. CBO has some influence, because people listen to the estimates as credible estimates, but the decisions lie with others. I have said that repeatedly to leaders of Congress in both parties. They would say, “Can we do this?” “Yes, of course you can do that. That is up to you, not me”.

I should declare an interest as an alumna of Harvard Kennedy School.

I have never heard that before.

Exactly, I thought I needed to. Professor Elmendorf, thank you for coming. The OBR forecasts are based on a five-year timeframe. My understanding is that the CBO uses a 10-year timeframe. Should the OBR consider using a more long-term forecast like CBO, and, if so, why?

I would be in favour of a longer-term forecast. One needs to recognise the growing uncertainty as one looks out, but there are a lot of policies worth considering that would have effects in years 6 through 10 that they will not have in years 1 through 5. It is always true that Governments can turn over and change policies, and we see a lot of that, but it is helpful for policymakers to have a slightly longer sense about what is likely to happen. In the United States we moved from five years to 10 years at a point at which the political consensus was to balance the Budget in seven years, because it was too hard, substantively and politically, to do in five. To do it in seven and to show you were doing it in seven, you have to have a projection that went out beyond year 7. That is the point at which we actually moved to 10-year projections. There is no magic to that. Your follow-up question made note that there are policies that would have positive or negative effects beyond 10 years, so there is no perfection in this, but it is, in my view, an improvement to look at a longer period.

Aghion and Howitt’s endogenous growth theory has shown that policies such as education, R&D and innovation drive productivity enhancements, but they take longer than five years, generally, to come in. Is that a major consideration here? Obviously, against that is also the five-year electoral cycle.

It is important in capturing policies that can affect economic growth, such as education, but also infrastructure investment, where there will be some effects by years 4 or 5, maybe, if you build something, but more effects if you look further out. There are tax policies as well. This is not a pro-spending position. If you reduce tax rates and there is more private investment, that can also have additional effects on growth as one looks further out. Also, policies that add to the debt can have more negative effects if you look further out, as the debt compounds. For all these channels looking at growth, a longer-term perspective can be helpful.

There can be advantages to having a longer period, because, if you make a forecast and the economy is not in balance, you need to have assessments of the output gap. If you have a longer period, you have a better chance of hitting that right than if you have only four or five years, especially if you had a big recession. That would be a reason why I would favour a longer-term view. If you were to look at some policies, especially with demographic features, you would need an even further horizon. That is why in the Netherlands we have the sustainability gap, which looks at whether your debt will be sustainable in the long run. In the Netherlands this has been quite helpful to consider things such as raising the pension age. You never do this for one or two years out, because then you have a lot of political hassle and very little to show for it, but, if you can show for the long term that it works, it is positive. Talking about a 10-year framework, I would be more hesitant about expecting to be able to programme the benefits of more education or infrastructure on productivity growth, because there the science is that it is likely, but it is not very precise. You have to underline that it is additional and not instead of or keeping the level there. In the Netherlands it is also quite popular to say, “I want to invest in education. Why don’t I get growth points for that?” but the science there is very limited. It does not mean that you should not do it, but you should not do it just because you think your growth projection will go up by 0.5.

What timeframes do you use in the CPB?

We use the Cabinet period of four years, but then we have a medium term, so that is eight. It used to be four, but now it is eight. We have this sustainability gap analysis, which goes for 50 years out. That kind of works in the same way as the OBR sustainability review, but there you have an enormous level of debt at the end, which people find difficult to get their head around, which I understand. In the Netherlands, it is translated to say that, if you want to make sure that every generation has more or less the same benefits, you have to adjust your fiscal position today with so much, or you can improve it or decrease it. That makes it more actionable. That has been quite useful in the Netherlands, as opposed to 300% of GDP debt where you think, “What should I do?” It translates it into what you could do and also gives policy possibilities to get this down, because in practice it always has to go down.

Can I focus in on that long-term forecasting and its value? It seems to me that the OBR in the UK does a very thorough job of assessing the long-term risks and liabilities, but it seems to be completely disconnected from the political discourse, probably to its frustration. It has suggested that we have an annual debate on it and that we should, in the Treasury Select Committee, initiate that debate. What advice or other ways can you give us, looking at our system and where we are now, to make that more meaningful as part of the political discussion? Is it down to us politicians, or what else can be done to make it a more meaningful intervention as a piece of data into our discourse?

Thank you for the question. It is a hard question to answer.

Yes, that is what I am here to do.

I am sure that that is why I am here. This really does depend. We have this expression: you can lead a horse to water, but you cannot make it drink. All that the OBR, or any fiscal institution, can do is to bring the information to the legislature and the public. Then it has to be picked up, or not. At different points in our history in the last few decades, the long-run issues have gained more attention or less attention. That is not because the CBO has done anything different. It is just because of the way the political debate has occurred. In your case, to centre something more in the policy discussion requires some policymakers to stand up and say, “This is a very important point to recognise”. The CBO puts out 30-year forecasts every year now with its 10-year projection. You want to be careful about the word “forecast”, because it is a scenario for the future. It is even less of a forecast than a five or 10-year number would be. Sometimes those get attention and sometimes they do not, so I do not think that there is any more that the OBR can do. It is a matter for you and your colleagues to say, “This is something that we think that Parliament and the Government should take seriously in their planning”. Then you can ask the OBR to play out some scenarios. Laura mentioned the idea of having policy options available, and CBO does the same thing. There is a set of policy options on the table. Most of those focus on 10 years, but some of them are about changing the pension age and so on. They go out over longer periods. You could have the OBR illustrate different ways of addressing a longer-term situation, but it will still come back to you and your colleagues drawing attention to that in the policymaking debate.

I thought you were going to say that. “It’s on us”—which it sort of is.

In the end, politicians or policymakers have to feel the need to act on it, so there I agree with Professor Elmendorf. The OBR could still improve on what is done. I think it now publishes the report in July, when everyone goes on holiday. That is not the best moment in time to get attention. The other thing is that some of the things can be made more actionable. That is what I said. In the Netherlands this is a part of the political debate. It became part of the political debate because there were analyses. There is always a group of high-level civil servants that produces advice on fiscal policy before the elections. They could come up with specific advice, and that was done because you do not have 300% of GDP debt but because there is something actionable: “You can solve the issue if you take measures for X% of GDP”. Translating difficult concepts to something you can act on is one thing they could do. The second is that they—either the OBR or the IFS—could put out scenarios, but you have to request for them to do it. When we suggested that during the review, they said that they did not feel that that was within their mandate.

That problem is quite key. It is this problem of defining a quite severe doomsday scenario without prescribing ameliorative policy interventions in the short term because of that apparent lack of licence to do so. Would it not be a helpful iteration of the OBR’s scope to say that, on the long term, there must be some clear preferences or consequential policy decisions that could be taken in order to initiate a more meaningful discussion among politicians and policymakers? Is that a viable enhancement?

That would be a useful route forward. In the Netherlands, we used the concept of comparable benefits for all generations, which sounds less political, because then you say you give every generation the same pleasure or not. That is the way that we translate this big chunk of debt into the action you should take. Then you can think about what the measures you could take are. This is only on demographics, I would add. You could say, “I will raise the pensionable age”, which is fair then for everyone in the future, or you could do something else—the other thing is of course healthcare spending—and then reduce that. Then you can say that every generation, or the next generation, has the same benefits in the future. That is the closest you could get to a neutral position. In the end, you need to bite the bullet, because raising the pensionable age is not an attractive measure, and savings in the NHS are probably not either.

If we were to summarise this chapter of evidence, it is summed up by Professor Elmendorf’s phrase: you can take a horse to water. The difference between the Netherlands system and the US system or our system is that not only is a horse taken to water, but it must drink as part of that pre-election and post-election coalition analysis, whereas here we might be viewed as shying away from choices. The outcome is that the Netherlands has debt to GDP of about 45%. We are about 100% and the US is north of 100%. As a mechanism for taking a horse to water and at least making it look at the water, do you think that there is merit in, for example, requiring the Government to respond to the OBR’s fiscal sustainability report with a detailed response of, say, 20 to 30 pages setting out their analysis of how we get public debt stabilised and, following that report, a proper one or two-day debate in the Commons about long-term fiscal sustainability? I do not think that we have had one since I got elected.

I am loath to be too firm on my recommendations for your political system, but I think there is a virtue, if you want more attention drawn to these long-term issues, in ensuring there is some official reaction to what the OBR says on the long run. Again, I want to endorse the idea of the OBR being asked by you all to look at alternative ways of addressing the long-term challenges in order to help elevate the issue and help people understand the sorts of changes that would be needed to produce greater stability over the long run. It is worth having a long-term discussion about that. In this country, when Paul Ryan was one of the leaders in the Congress and he was very committed to making long-term changes in programmes, he brought that issue to the fore. Since then, neither of our political parties has been that interested in addressing these issues. I think that they could be encouraged to do so through the kind of mechanism you described.

I would say, “Why not try it?” I must say that, in the Netherlands, we have a tendency to discuss and debate only on the Budget, never the accountability at the end. We wanted to have more attention for that, so we introduced a new day for that, with a debate, and in the end it kind of fizzled out. It is not a guarantee that, if you say, “Come up with a response”, and have a date and arrange for a debate, it will work, but it could be the ignition of something new, so why not try it?

To pick you up on that, you are suggesting as well a proper report and analysis of how the Government’s promises to get debt down or achieve a greater fiscal sustainability have actually out-turned in practice. Focus on that too, so that there is an element of accountability after a certain number of years that is formalised in some way.

There, it did not work, I would say. If you want to have a debate, you need a response by the Government. There is a response. I read that there are always five pages of a superficial, or at least experienced as superficial, response. If you want a debate, you probably need to have a focused question and define what type of answer you would expect from the Government, for instance stating, “We have this debt. How do we get this under control and with what types of measures?” Then you could ask for options. If the Government do not want to contemplate new policies, you could at least ask them to come up with scenarios on how to address it and then you have a debate on the scenarios. You get more enrichment of the debate on how to move. A Government who do not want to address the issue will never come up with a specific answer. If you say, “We want to explore the issue. We want to have options of how to address it”, while not asking them formally to do it, that is easier.

That is very helpful. I just want a quick answer to this. It strikes me that you would have the publication of the report and the government response, which would set out scenarios to make the fiscal position sustainable in the long run. Perhaps it is quite attractive to have the OBR mark those scenarios, at least to say, “That is credible. That is not credible”, so that there is some sort of anchoring to the debate about long-term fiscal sustainability. The problem I face is that I do not think that our fiscal situation is sustainable in the long run. We have to face into this and we are not facing into it. What about that as an idea?

That is a very interesting idea. You are, of course, better versed on your political system than I am, but I thought that there was a hesitance on the part of the OBR to come up with its own scenarios or policy options. If you asked the Government to come up with the options, not requiring them to take the options or make the policy, and then asked the OBR to test or cost them, that would be a proper delineation of the various responsibilities. Perhaps it might be comfortable with that. That would be an interesting way forward.

Yes, that is a reasonable way to proceed. If you do so, I will be curious to see how it works.

I would like to start by asking Professor Elmendorf how the interaction works between the Congressional Budget Office and the Executive. One thing that the Office for Budget Responsibility has here is the privileged access to the Treasury, what it is planning to do, what the measures are, what the assumptions are and what the tax revenues are. I wanted to see how that interaction works in the States.

There is a lot of interaction at the staff level. When I was CBO director, my analyst who worked on tax policy would talk with civil servants who worked on tax policy at the Treasury Department. Those who worked on various spending programmes would talk with people in the executive branch’s budget office or in the various Departments of Government. The important thing is that those discussions were all kept at a staff level. I never spoke myself with the head of the executive branch’s budget office, or with a political appointee at the Treasury Department, or anywhere else in the Government. The idea was to make sure that this was just a discussion among technical people on the way they are seeing the incoming data or whatever, and not something where there is a risk of political pressure being applied.

Do your staff have the resources to come up with their own tax revenue projections?

Yes, and we view that as crucial here. I understand that that is not the case for the OBR. It is part of establishing the independence of the Congressional Budget Office. That independence is important for credibility, and it is important in both fact and perception. It is crucial for these institutions to not only be objective and unbiased, but also be seen as objective and unbiased. The higher the walls you can build, as long as there is some flow of technical information, the more that credibility is reinforced. I would support the OBR receiving the resources to do its own estimates of what tax revenue will be, not that they would necessarily be better, but that they would be more clearly separate from the Government’s view.

Would they be quite substantial resources that you have allocated to that?

They would be substantial resources. At the Congressional Budget Office there are maybe 10 people part of whose job is to do the revenue baseline, as we say, so the basic revenue forecasts, and a larger group of people who look at policy changes. It is noticeable, but it would not require, I think, a dramatic change in the scale of the OBR.

In terms of the independence of the Office for Budget Responsibility, would you say that that independence of the resources that it has is a crucial question? How would you strengthen the independence of the resources that the Office for Budget Responsibility currently has from the Executive?

Do you mean the resources in terms of money, or in terms of the models that you just discussed?

I mean the resources that it uses to either employ people or develop models, so the things that it uses to produce its output.

At present, the OBR has a budget. Independence of an organisation depends on having enough resources to do the work you need to do. The bigger risk is not firing the head of the independent fiscal institute, because that is very visible, but having limited or too few resources. That is the way to kill an organisation, because then, in the end, it cannot do the work, it loses reputation and everything is done. During the external review we discussed the size of the resources and, by and large, that seemed all right. We always said that, if there are new tasks, you should get new resources for that. If you want to have the costing of these sustainability measures, for instance, that will probably require more work. We also noted that, while we were there, its budget was one year out, but that has been addressed by the present Chancellor, I think, so the resources are okay. When I came, I was surprised with the way that the modelling was distributed across the various parties or parts of government. In the Netherlands, we make our own macro forecasts. That is what the OBR does as well. We also make our own tax revenue forecasts, but, I must admit, if there are individual tax measures that need to be costed in the Netherlands, we also rely on the Government for those. Then we do a kind of certification or test of reasonableness of those measures. If you want to do that, you really need far more people. When we looked at the OBR, we thought that model maintenance was at risk, because urgent things always take priority over important things. That is why we highlighted in the external review that you should do more on model maintenance and make it a plan, so you also are accountable if you are not doing it.

If a Chancellor wanted to reduce the number of resources at the Office for Budget Responsibility by cutting the budget, what checks and balances could there be in the system to prevent that from happening?

When we did the third review, we said that, if the Chancellor wanted to consider that, they should inform your Committee on that. We were wondering how to deal with this and we asked around what would be fitting within the UK institutional framework, because every country is different. People advised this, because this is also the way that you deal with it if you want to fire the chair of the OBR. It has to go through you. We said that, if there are major cuts in the budget, you should at least signal it to the Treasury Select Committee, because we were told that that would be the best way forward, as opposed to alternatives that had been reviewed. I must say that we are not experts. We discussed it with, I think, one or two people and then we thought that that would be the best way forward. That was the reason why we put it into the report.

The OBR will only score supply-side measures if they increase or decrease potential output by 0.1% by the fifth year of the forecast. Do you think that that is reasonable? What do you do in the Netherlands?

In the Netherlands, like the OBR, we take on Keynesian effects for demand side. We take on supply effects in the tax that are linked to labour supply, because there you have lots of expertise and evidence of what the size would and should be. These are parts of the dynamic effects that the OBR is also undertaking. We were hesitant and reluctant to do that on the productivity side, because there evidence is very thin on the ground. I think that it was a compliment to the OBR that it was willing to investigate this area. That is also what we said: “You are not conservative. You are willing to go out and look at new things”. That is what we said, but we also thought that it was not cautious in that sense. That is why we said in the report that it should review what it had done. I think this review came out in November 2025 and it struck the right balance there. If there is no evidence, you should not go for it. If it is very limited, so 0.00006, you should ignore it, because then, compared to the average uncertainty around the figure anyway, it does not really make sense. I thought that, yes, it is good to still do this type of exercise. If it is big enough, if it is a comprehensive, logical package and if there is evidence, you should do it, but you should not overstate the options. The report of November suggested that it was good to have this assessment and to think again about whether you should go after everything or just look at the big chunks.

I will make a few points to add to what Laura said. One is that the evidence base for this kind of analysis is growing. The Congressional Budget Office now incorporates more of these sorts of dynamic effects than it did before. The most recent forecast of productivity growth talks about the role of immigration, which has reduced in the United States. That is reducing the inflow of inventive people. It talks about the role of higher tariffs making production less efficient than it would be otherwise. It talks about the role of education and of artificial intelligence. There is not a lot of evidence for those things, but in most of those cases there is now enough evidence to be pretty sure what the sign of the effect would be, if not the precise magnitude. There is value in signalling that, but it is also true that you need to have some sort of consistency across different kinds of estimates, or it will look like the organisation is playing favourites and favouring some things over others. In the Congressional Budget Office’s analysis of policies, it has mostly operated with a threshold, where policies that are big, so above that threshold, get a dynamic analysis, as the big tax cut and spending cut Bill of 2025 did, and policies that are smaller do not get that analysis. That creates its own inconsistency though, because it means that, if you have a policy this big for infrastructure and a policy that big for infrastructure, one might be called a zero and one is called something bigger. There is no perfect way out of this, but it is important to build the tools. Another thing I would say is that this is a case where scenarios and options can be useful. Even if a budget agency does not have the time, resources or confidence to estimate the effects of a lot of different specific policies, it can write a report or say, “There is evidence that investment in research and development spurs productivity growth”. It is a way of signalling and publicising that certain kinds of policies can be good for growth in the long run, even if you do not have enough confidence, time or staff to do that for specific policies as they come up.

You spoke about thresholds. Is there a similar concept in the CBO to the 0.1% we are seeing here in the OBR?

At different points, the House of Representatives has established rules for its own behaviour that say that, if a piece of legislation is estimated by CBO to have an effect above a threshold on the Budget, without taking account of these sorts of dynamic effects on economic growth, that Bill looks big and important enough that the House wants CBO to take the time to do that kind of dynamic estimate. That is what happened last summer. That very large Bill was well above the threshold and therefore CBO was instructed, through that rule. Its mandate, in essence, was brought in for that Bill and it produced a dynamic analysis that captured the growth effects, last summer, as the House of Representatives was debating the Bill.

There is one final question from me for you both. Should the Treasury be braver in disagreeing with the OBR on growth forecasts? In the UK, as I am sure in the Netherlands, it gets castigated in the media if it disagrees with the OBR. The OBR is not God, as Professor Elmendorf said earlier.

Of course, people can always try to disagree with forecasts. If you disagree with a macro forecast, you need to substantiate it. You can always discuss whether you disagree with the fiscal forecast, where I think Governments should have more leeway. To be honest, if the CPB were to make a fiscal forecast, we would not take the DEL figures of the Treasury at face value. We would take other figures and end up higher or with a worse deficit. In the Netherlands, there are differences. The CPB has figures and the Treasury has figures, and there are differences. These differences are accepted, but they are always limited and never on things such as whether you run out of whatever your version is of fiscal headroom. In the Netherlands, people believe the CPB, for the right reasons.

I agree with Laura. Disagreement can be good, but one should bring evidence. The Government should recognise that it will, in general, be viewed somewhat sceptically about its optimism. Last summer in this country, CBO said that the big Bill would have a very small effect on economic growth. The Administration said, “No, it will have a huge effect on economic growth”. At least three outside organisations, university or think-tank-based organisations that do fiscal estimates, came up with their own estimates, and they were all tightly grouped near CBO’s. CBO and these three independent organisations were all down here, and the Government were way up here. That did not do anything for their credibility. It did not move many people into thinking they were right. I am all for disagreement, but the Government need to recognise that, without having a strong evidence base and building credibility, they are not going to move the discussion very much.

I just wanted to pick up on the point that Professor Elmendorf raised about assessing legislation that is in process. We have had a situation in the last couple of years where the Planning and Infrastructure Bill got scored by the Office for Budget Responsibility whereas the Employment Rights Bill did not. Although they were both taking their time through Parliament, one was deemed to be credible enough and near enough to completion to be scored and the other was not. Do you apply a consistent rule to legislation in flight? If so, what is it?

By law, CBO is supposed to produce an estimate of every Bill that is voted out of a Committee of the House or Senate. Sometimes, the Congress moves so fast that the estimate does not come out in time, but that is unusual. So that is consistently the case. In a given estimate, CBO may or may not be able to take account of everything that one can think of, as we have discussed, but every piece of legislation is supposed to get an estimate after Committee and before the House or Senate as a whole would vote on the legislation.

That is set in law.

That is set in the founding statute for CBO, in fact. It has been roughly honoured. Not everybody cares what CBO’s estimate is—that is their prerogative—but there is almost always an estimate produced.

Thank you. That is very helpful.

Partly, it is a relentless commitment to non-partisanship for 50 years. We never took that for granted. It has been a high concern of every CBO director to act in a way that both is and appears to be fully independent. When CBO was created, the first director was appointed by Democrats who were in control of the Congress. A Democratic President was elected and the CBO issued what was viewed as a very negative review of President Carter’s proposals. People in the Congress were surprised. “Oh, I guess they mean it when they say independence”. We kept that going. I felt that it was something we had to earn every day. Every time I testified and in every report we wrote, we had to be very careful to show that we were doing our best and an honest job.

In the Netherlands, the CPB was established in 1945. We are old. The first director was Jan Tinbergen, the Nobel prize winner. That has a kind of stature in and of itself. Of course, as Professor Elmendorf said, you need to be conscious of why you do the things that you do. You have to explain it and be sure that it is not partial. If we take a figure that is different from the figure that the Government would produce, we have to explain why we do it. I will take a recent example in the Netherlands that is kind of a different one. The Dutch Government were very ambitious in raising defence spending. They put in reserve money. The CPB thought, “They are never going to be able to spend it because they are not going to be able to find the people or to purchase the stuff that they like to purchase in the period that they have signalled”. They put it in a lower figure and they explained why that was the case. People could then assess whether that was reasonable or not. The other way around, if the Government said, “We will keep primary education spending constant even though the number of pupils is rising”, they would question it or at least signal that, if you were to continue on this basis, the number of pupils per teacher would rise from 10 to 20. You would make visible the implicit pressure that would then be contained. In healthcare, “If you do this, waiting lists will increase and increase”. The way forward is to look at the figures, study whether they are proper and explain if you do not feel comfortable with them. You have to be very transparent about it. The OBR is very transparent in what it does. It is far more transparent than the CPB ever was. It is not that it does not want to be transparent, but it does not feel like it has room to manoeuvre. In the external review, we suggested having a standstill version or scenario. They feel uncomfortable with that. The Irish do it. In the Netherlands, whether this is political or not would not be a topic of debate. It is basically stating, “If you want to keep the same arrangement, you need more money because things have changed”.

Thank you. That is very helpful. The final question is from me. We had a lot of speculation swirling around our Budget last year. There is some suggestion that, if the OBR was able to publish its pre-measures forecast earlier than the Budget, it might have stemmed some of the speculation. Do you have any quick thoughts on that?

I support releasing the pre-measure forecast in part because that is what CBO does. The logic that we have applies to you as well. Policymakers need to understand what is likely to occur in the economy and the Budget if they do not take any action and then see the difference that might arise if they took particular kinds of action. The logic of this work is to say, “Here is what happens if policymakers go home and here is what they might do”. To do that, you want to have a pre-measure forecast released. You want to see what happens as various measures are proposed to add to or subtract from those outcomes.

I would support that as well. We do this in the Netherlands. Twice a year, we have a forecast. We have one in March, which is the starting point of the budgetary discussions in the Netherlands. In September, you have a second one. It used to be the case that it stayed secret until the parliamentary opening, but there was a lot of leakage and releases, and we thought that was not proper. In my term, the CPB decided that we would publish the pre-measure forecast in August. It was a comparable problem to what your Government apparently experienced. There was a secret pre-measure forecast, but then people all talked about it. In the Netherlands, it is quite typical to have a coalition agreement. It is a different kettle of fish. The Opposition were concerned that there was no level playing field and we sympathised with them. Because the leakage came from Government, we have a policy of, “If we discuss things with you and they leak, we immediately publish them”. This is one way to make sure they do not leak. If you do leak, you get punished. By way of punishment, we just publish it. Because this was a pattern that happened every time, we changed our policy rule and we said, “We will also publish a pre-measure estimate in August”. We learned this the hard way because there was a lot of debate on whether we shared information earlier with the Government than the rest of Parliament or the public at large.

Thank you. That is very clear. Can I thank both witnesses very much indeed for joining us? That is Professor Douglas Elmendorf and Laura van Geest. We will be producing a report on this in the near future. We will make sure you see a copy of that. Thank you for your very useful and pertinent evidence. Witnesses: Tom Josephs and Professor David Miles.

Welcome back to the Treasury Select Committee on 19 May 2026 to continue our examination of the Office for Budget Responsibility 15 years on. I am delighted to welcome—they are regular visitors to this Committee—Professor David Miles, who is a member of the Budget Responsibility Committee, and Tom Josephs, who is also a member of the Budget Responsibility Committee. Normally, it is three, but the recruitment for the chair is under way at the moment. I just wanted to check whether we have any update on that timetable. Who wants to take that?

First of all, thank you very much for the opportunity to give evidence on this inquiry, which is very important for us. The scrutiny that this Committee provides to the OBR is really important for us. We very much look forward to seeing your report. I am afraid I do not have an update. It is a process that is run by the Treasury. It is really for the Treasury to update.

That is fine. Thank you very much indeed.

It is very good to have you here. Could you start by summarising, each of you, what the OBR has done well over the last 15 years and where it can improve?

The remit of the OBR sets out that we are here to assess the sustainability of the public finances. That is clearly set in our legislation. We do that by producing forecasts, costing government policies, producing analyses of risks and the long-term sustainability of the public finances, and assessing whether the Government meet the fiscal rules. The big difference that the OBR has made in those areas is twofold. First, as you heard from your previous experts, we produce all our forecasts, costings and analyses in an independent, unbiased, evidence-based and objective way. That allows for better quality analysis for policymakers to take their decisions on, and allows Parliament and the public to scrutinise those decisions. Secondly, the really big difference that the OBR has made is to bring a real step change in transparency on the public finances in the UK compared to the information that was published prior to the OBR. The OBR provides a huge wealth of information that was just not available previously, both in our central forecasts and policy costings, and, really importantly, on the risks and uncertainties around those forecasts and costings, but also on the longer-term sustainability of the public finances. I will just add one thing before I pass to David. The ability for us to be transparent in that way and provide Parliament with all the information that we do really stems from the model where the OBR is producing the forecast and costings that the Government use in their Budget-making. That gives us access to government information that we would not get if we were the type of fiscal council that sits on the sidelines and comments on the Government’s own forecasts and policy costings. It is that access to government information that allows us to be as transparent as we are.

Just to add to that, the key thing that the OBR has done, with some success, is to maintain its impartiality and be viewed as making cool, evidence-based assessments of the fiscal and economic outlook. That is not the same as producing central forecasts that always closely coincide with what happens. The impartiality is important and is linked to transparency because, without transparency on what the OBR has done, it is much more difficult to judge whether it is impartial or not. If one looks back to the period before the OBR, we got into a situation where at the time of the Budget the Treasury would produce a very thick book, the Red Book, and it came to be perceived, with some justification, as a bit more of a PR exercise. If you were being very judgmental, you might say it was a bit more like Soviet propaganda than an assessment of what was likely to happen on some central forecast. It has been an improvement to have an institution that will not necessarily produce forecasts that are much more accurate than other people but will do so in a very transparent and unbiased way, using information from the Government, sometimes with a sceptical eye on what is fed to it. In terms of things that one could do better, the decision a few years ago to more explicitly judge potential impacts and indirect effects of government policy on productivity, output and the economy, for either good or bad, has been a step forward. One would have been open to legitimate criticism if one had just said, “When we look at a government policy, we just ask how much more tax revenue it brings in or how much it is going to cost, if it is a tax or spending measure, and leave it at that”. We would have been open to the criticism that, when Governments take decisions that are likely to boost productivity and output, and bring in more revenue, we are just ignoring that, which is an asymmetric way of thinking about government policy. That would have been a fair criticism. We have gone in the direction of—some people call it dynamic scoring—assessing both the pluses and minuses on the wider economy of policies. It has definitely been a case of pluses and minuses. To give you an example on the minus side, the decision, understandable in many ways, to freeze thresholds to bring in more tax revenue, which under current proposals will last, from when it started, the best part of 10 years, comes with a cost. It is a big increase in the income taxation of households. It causes changes to the incentive to work. We have put in estimates of, in a sense, the damage that it does. It is very difficult to raise lots of taxes without doing some damage. That is part of dynamic scoring, as much as the positive things, such as when Government decide to spend more on public sector investment, planning measures or some of the other things where we have taken a positive view.

That is very useful. There is quite a lot to unpack there, and I am sure colleagues will do. One thing that has been said is that the OBR has been a bit of an outlier by being too positive about UK growth prospects over a prolonged period. The Institute for Government, for instance, has claimed that you have had an optimism bias around productivity, which has driven overestimates around growth. Would you agree with that verdict and that some change has been needed?

There have been five occasions in a relatively short period, since 2010, when the OBR has reduced its estimate of what you might call the expectation of future trend productivity growth. It started out back in 2010 as an assessment that the UK would bounce back to its long-run average rate of growth productivity of about 2% or slightly more than 2%. Six or seven years later, when it became obvious that, after the recovery from the financial crisis, the UK economy did not seem to be getting back to the same trajectory it had been on before, the 2.2% became, from memory, 1.7%. A bit later it became 1.5% and then it was reduced again to 1.3%. The judgment that we made at the back end of last year was that 1.3% was probably still on the optimistic side and we have come down to 1%. It made sense to wait until the evidence became a bit clearer each time that judgment was made that, unfortunately, the UK did not seem to be getting back on the trajectory that we thought before that. The latest decision back in November of last year was made after we thought enough of the fog had been lifted, if you like, after the very detrimental short-term impacts of, first of all, covid and then the huge increase in energy prices after the Russian invasion of Ukraine, as well as the real difficulty in judging the situation when, if you are looking at productivity, it is output relative to employment and the employment picture was very foggy because of the problems that the ONS had with its Labour Force Survey. There is always a good case for not rushing into a judgment, particularly one that is going to have a material impact on the fiscal outlook. Others do not face quite the same implications, if you put it that way, for private sector forecasters or for think-tanks. People will pay attention to when they change their view on productivity, but it will perhaps not have the same implications as when the OBR does it. There is an argument for waiting until more evidence is in that might trigger changes in forecasts by, for example, City economists.

I will just add one thing on that. Going back to the point on transparency, on that judgment, clearly, there is a huge amount of uncertainty. There always has been since the judgment was first made back in 2010 on the post-financial crisis outlook. The OBR has always acknowledged that uncertainty and provided scenarios around the central forecast of productivity to show the potential fiscal impact of outcomes that are worse or indeed better than our central forecast. Every year, we have also done a very detailed evaluation of the performance of our forecasts. That allows us to spot where we are serially getting things wrong and take action to address those. As David says, we have done that on productivity. When we did that in November, we published a very detailed paper on why we were taking the judgment to downgrade productivity. It is both the judgments that we make and the transparency that we provide around the risks and the performance of those judgments over time that hopefully allows, again, for better policymaking and better scrutiny of policy.

Professor Miles, is the whole debate about the public finances too precise? Are we spending too much time on single-billion figures rather than perhaps greyer or less precise general trends?

Yes, I have a lot of sympathy with that view. We have got ourselves into a slightly tricky situation in the UK. Looking at the central forecast for where the current Budget will be four or five years down the road—it is coming closer now with the move to a three-year horizon—is a bit of a zero-one or pass-fail story. The reality is that, if you go from small positive headroom of, say, plus £5 billion to a central estimate of minus £5 billion, that is really a rather trivial move in a world in which government spending is £1.5 trillion. The probability—this is something that we try to emphasise in our work—of being on the right side of a government target, if headroom is plus £5 billion, might be 51%, and, for minus £5 billion, it might be 49%. Crossing that line of zero does not deserve to be judged as passed with flying colours or complete failure. There is more that the OBR can do—and we will try—to emphasise the broader context of the fiscal situation. That is partly by emphasising probabilities and what it means to have a plus or a minus number on a central forecast for headroom. Almost inevitably the central forecast will turn out to be rather far from the out-turn three or four years down the road. At the same time, the legislation does say that the OBR shall provide an assessment of whether it is more likely than not that the Government meet their fiscal targets. That is why I can understand that the media focus very much on that single number. Because the media is going to focus on it, particularly in a difficult fiscal situation, the Government can hardly ignore it and have to pay a great deal of attention to it themselves. In some ways, we would all be better off to focus on a more nuanced and broader picture of the fiscal outlook than just that single number.

How might that nuanced approach relate to reality?

As David says, we have tried to provide more information over time on the wider fiscal health of the UK. For example, we did that at the spring forecast this year because the Chancellor decided that that would not be a policy event. In a sense, that gave us more space to provide that wider analysis. For example, the IFS suggested in its report using a traffic light analysis of wider fiscal indicators. We included something similar to that in the report. We provided more information on how the UK fiscal picture compares internationally and more on the history of the public finances in the UK, how we have got to where we are now and recent out-turn as well as the medium-term forecast. Understandably, that slightly got overshadowed by the fact that the conflict in the middle east broke out at pretty much the same time, but that is something we would hope to build on in the future. There is also the question that you were discussing with your previous panel about how we can draw more attention to the long-run sustainability analysis that we produce, which, again, is an important wider source of analysis on the public finances. As David says, the headroom number is also important. The Government have set out very clearly their fiscal rules. Given that, as we all know, we face a very challenging fiscal situation, sticking to those fiscal rules is seen as very important to credibility. The headroom number gives an indication of how much fiscal space the Government have against those fiscal rules. This is not to say that is not also an important part of the overall picture, but I agree that it should be seen as one indicator among many.

How do you respond to criticism from the IPPR that the OBR does not adequately capture the second-round dynamic impacts of public investment, such as where public capital crowds in private investment?

We do take that into account. In the autumn 2024 Budget, there was an announcement about a big increase in public sector investment, as you will know. Our assessment was, if I remember correctly, that for every £1 of extra public sector investment there would be £1.30 total investment because you get another 30% on top of that as the knock-on effect of private investment being crowded in. I do not think it is quite right to say—I know you were not saying it, but others have said it—that we do not factor in helpful second-round effects when we look at higher public sector investment. Of course, there are potentially negative second-round effects sometimes. If there was a huge increase in government spending on investment at a time when the economy was right up against its productive potential limits, that might crowd out some private sector spending. It might crowd out largely consumption spending, but there could be a bit of a crowding out of private sector investment. At the time of the autumn 2024 assessment, it was that there was substantial net crowding-in of private sector investment.

Of course, crowding-in can be much higher than 30%.

Yes, it can be. We based that on the approach that we use to think about the positive effects of public sector investment on productive potential. We used the same framework: production functions, profitability of investment, what it does to capital-labour ratios, wages and all that machinery. You can use that to work out how much of a boost there might be to private sector investment. We used the same framework as we did to judge the first-round direct effects of public sector investment.

Has it ever been the case that the OBR’s threshold for scoring growth policy has resulted in the Government delaying or even blocking investment decisions because it does not get the benefit of them in the OBR forecasts?

The threshold that we have set to at least consider policies is that they should have an expectation of at least a 0.1% impact on GDP, one way or the other, five years down the road. In a way, that is a really low threshold. An impact of 0.1% GDP five years down the road is pretty small. That threshold has not ruled out great swathes of policies that the Government might otherwise have done. It is always going to be the case that the Government are going to be frustrated when we make any judgment that a policy does not look like, on balance, it is going to move the dial that much at five years, particularly in a difficult fiscal situation where you have to pay for the investment up front. We get that coming back to us, of course, in meetings with Treasury, the Chancellor and advisers to the Chancellor. It is inevitable that there is sometimes going to be some frustration that policies do not get scored more favourably. The reality is that, if there was no hint of that whatsoever, we would probably not be making unbiased judgments about the impacts of policies.

You might have a number of policies that individually are below 0.1%, but in aggregate—

Yes, if they form a coherent whole, as opposed to saying, “We have these five policies. They are all in different areas. If you add them all together, they just get above the 0.1% threshold”. That is gaming the system slightly. If they clearly are coherently part of a package, it is a perfectly legitimate thing to do, and we would do that.

On the particular question on public investments, our estimate in 2024 was that that would have an impact of plus 1.4% over five years. We did not have that threshold in place at the time, but the increase in public investment at that point would have been above the threshold so we would have incorporated it into the numbers. We also had analysis that showed that, if it was sustained over the longer term, the benefit potentially would increase. The other thing on the actual analysis that we did on the question of crowding-in is that we reached our estimate on the basis of a very thorough study, looking at the whole of the economic literature internationally and reaching a judgment based on that. Of course, there are upside and downside risks to that, but our job is to do an evidence-based objective assessment—that is what we did—of the economic literature as it is.

Finally from me, we have discussed a lot about the five-year horizon. We have heard that the CBO uses a 10-year horizon and the CPB in the Netherlands uses a four and eight-year horizon. Is the five-year horizon too short? Is it correct? If it is correct, where do you disagree with the Netherlands and the US?

To start with, it is probably worth clarifying that the forecast horizon for the Budget forecast is set by the Chancellor. If the Chancellor decided to set a different forecast horizon, we would produce that. We also produce longer-term forecasts in the sustainability report and on particular policies. Again, the Chancellor asked for that and changed our legislation to allow us to do more on things such as planning and investment where there is a longer-term impact. We have done an analysis of that in our reports. In terms of the forecast horizon for Budgets, there are pros and cons of moving to a longer period. As people have testified to you, it is the case that some policies will have more of a material impact over a longer timeframe. Having a longer timeframe potentially captures more of that, although, again, it is very uncertain. There are risks around that as well. It would require Governments to set policy over 10 years because we have to produce a forecast that is based on government policy. Therefore, Government would need to say what they are going to do on tax thresholds and departmental spending limits over a 10-year period, which spans a number of Parliaments. There is a question as to how Governments would do that and whether they would be happy to. It is also the case that, as you go further out, your economic and fiscal forecast just gets much more uncertain and the information value of it diminishes. Going back to the policy point, policy commitments made on a 10-year basis may also be seen as less credible.

Sure, but you get the net present value today of a policy leading out for five to 10 years in terms of gilt yields and so forth. Is there an argument that, if you were scoring them from years five to 10, bringing through policy changes that would potentially impact gilt yields might improve the Government’s fiscal position today?

As I say, there are certainly some policies—we have set out analysis of this—that are likely, if they are sustained over a 10-year period, to have increasing benefits in terms of the impact on the economy and therefore the fiscal position. Again, the question is the degree to which those commitments would be seen as credible over that period and therefore influence gilt pricing in the short-term.

Is the question whether we think the Treasury should produce the forecast rather than the OBR?

As you have heard from previous witnesses, the OBR is quite unique among independent fiscal bodies around the world in producing the forecast that the Government use to produce the Budget. A lot of fiscal councils, in essence, assess the Government’s own forecast, make recommendations on policy or assess whether the Government’s policy decisions are the correct ones from a fiscal sustainability point of view. As I said earlier, the key benefit of the OBR doing the forecast in the UK context is, first, that we do it in an unbiased and objective way and do our best to remove any optimism bias that may be present within Government from those forecasts and analysis, including the policy costings. That leads to better analysis for the Government and better analysis for Parliament to scrutinise policy decisions. It also means we are able to provide the transparency that I talked about in the answer to the first question. If you are not inside the machine, so to speak, producing the numbers that the Government use, and therefore the Government have to provide you with the information in order to do that, you would not be able to provide the transparency that we do now. I go to an OECD meeting where all the fiscal bodies from around the world are present. The No. 1 concern and difficulty that they face, if they are commenting on rather than producing the Government’s forecast, is that they cannot get access to information and they are not really listened to, frankly.

There is nothing that stops the Treasury from analysing any measure that a Chancellor might adopt and giving a fair assessment of it to the Chancellor. Where one might have a problem—I am not saying this is a reality of exactly where we are today—is that it would be a bad world in which all those smart people in the Treasury directed all their efforts at trying to change the OBR’s mind on something rather than telling the Chancellor, “This is what we think is the best thing to do”. That is a question for the Treasury. There is a danger that one has moved a little way down that road. The other thing that I would say is that it would be a very healthy world in which a Chancellor could say, “The OBR has come up with its own forecast. It has made its own judgment on these things. I respect all that, but I think it is being a bit too pessimistic on this. We are going to do this, even though it does not get counted in the OBR. It has a more pessimistic outlook on the fiscal situation a few years down the road than we do. We will see who is right, but I think it is the right thing to do, and I am going to do it”. It would be excellent if that were to happen more often than it does.

Yes, exactly.

Just very quickly, I am sorry if I misunderstood your question originally, but I have two things to add. First, on policy costings, the Government are responsible for producing those costings already. That is all done by government analysts. Our role is to assess whether we think those costings are reasonable. On supply-side policy we take the lead, but we still work very closely with government analysts on those policies. The exchange that you are talking about is largely a healthy debate. There is a lot of expertise within the Treasury and other Departments on these policies, as you would expect. We are very open to listening to that, but we then take an independent view. I would add that part of the reason that we introduced the 0.1% threshold for scoring supply-side measures was that we discovered that we were using up a lot of resource looking at quite small policies, which had a very small or uncertain effect. Quite often, if those policies ended up not being delivered, or not being delivered as planned, the effects would fall away to virtually zero. Part of the reason for introducing the threshold was so that both we and the Government could focus more on those policies that are likely to have a more material impact.

It is not all lobbying. Lobbying suggests that it is one-sided and somewhat partial. It does sometimes seem that way, but that is not the norm. The norm is that we get information from all kinds of Government Departments. It is not just the Treasury. It is really quite impressive most of the time. We push and prod a bit, but I would not want to give the impression that anything that we get from any Government Department is optimistic, biased or lobbying. It does not feel like that from where we sit.

I do not think so. The fiscal framework has the great merit of being very precise and very clear. There are two main elements to it, as you well know. “Here is where we think the current Budget might be in a few years’ time”. Being on the right side or the wrong side of the Government’s target is whether or not the current balance is in surplus versus being in deficit. There is a separate rule on the debt-to-GDP ratio. In a way, it has the great merit of being clear to understand. We can produce a lot of information on it. It is pretty clear from our analysis whether you are on the right side of it or not. There is a separate issue about whether it is a broad enough picture of the fiscal position, particularly if you look into the longer term, but, in terms of how this works at the minute, there is substantial clarity about what the rules of the game are. I do not think there is anything in the institutional structure that stands in the way of the OBR making a judgment about the fiscal rules.

Professor Miles earlier mentioned Soviet forecasts. Turning to the fiscal risks and sustainability report, you will have heard my questioning earlier, which was really aimed at getting some glasnost and a potential menu of perestroika for the long-term sustainability of the public finances. In your long-term forecast, is there a role for a greater number of scenarios or recommendations to put the public finances on a more sustainable path? As an example, the out-turn is very different if tax thresholds increase at average earnings as opposed to CPI or are in fact frozen.

Certainly, it is very important to provide multiple scenarios in that report. Clearly, you are doing 50-year projections. There is a huge amount of uncertainty. These are not forecasts. They are different scenarios for how the economy and public finances would evolve under different assumptions around demographics, policy, etc. We try to have multiple scenarios.

Do you make policy recommendations or choices?

On policy, it is very clear in our legislation that we should not make any kind of recommendations on policy. That is an important part of the OBR’s governance structure and one that we are very careful to stick to. Through our risk analysis, though, we do present the implications of different kinds of policy mechanisms for the long-term public finances. As an example, in last year’s report we did a deep dive into pensions and we very clearly there showed the implications of the triple lock on the outlook for the public finances. We showed the implications of different choices on the state pension age. Similarly, we did a deep dive on health spending. We showed the implications of better working-life health for the public finances through the impact on spending, welfare and tax revenues. Although we do not make policy recommendations—I think that is important—we can clearly draw attention to the things that matter over the long term.

The Government provided a pretty short response to this report. Do you see merit in a longer and more comprehensive response to your very good report, which involves a lot of hard work?

Certainly, we are always looking for opportunities to raise the profile of that work. We think it is really important. It is one of the most important reports that we produce. It gets a lot of attention from this Committee and the Lords Committee, and it does get quite a fair amount of attention from the external analytical community. Finding ways to get more traction through the policy process would definitely be a good thing. In the past, the Government have produced a longer response. There was one year where they produced a very long, 100-page response. At the moment, that is not the case. As you said, there is not a parliamentary debate on the report. From our perspective, raising the profile of it and getting more policy focus on it we would see as a good thing, yes.

Could the Government response itself be marked in some way by the OBR? What if the Government were required to set out what they believe are sensible measures to ensure that the public finances are sustainable in the long run, and the OBR then got to mark or comment on them in some way? That would not necessarily be a pass-fail or a numeric number, but quite simply a view on whether it is credible or incredible.

Again, that would be a pretty big step away from the OBR’s basic set-up in legislation, which is that we do not make normative commentary or recommendations on policy. There is a danger that that would draw us too much into the politics of this. Our role currently is very well defined as an analytical organisation that produces the analysis. It is for others to make policy decisions and scrutinise those decisions, rather than for us.

What about a two-day or a one-day debate on fiscal sustainability following your report and that becoming mandatory?

That is for Parliament to decide.

I know; I know.

As I said, from our perspective, we would definitely see it as important to get more focus on that report.

I have one final question, then. In my humble opinion as a politician, we have a long-term sustainability of public finances issue that we do need to start facing up to. You are very carefully telling me that this is my problem and not your problem, and you are quite right to do so, but we do have this problem. The last session was instructive on the outcomes for the Netherlands, us and the United States, although at least we do not have an annual budget shutdown drama. Is there anything that we should change on our side, other than the menu of things I have mentioned, to put long-term sustainability firmly in the public and political eye?

It is a huge issue. Ever since the OBR has been doing long-term projections, it has pretty much always painted the picture that, under what you might call current policies, it might look all right for five, 10 or 15 years, but beyond that the debt just starts rising in a way that is clearly unsustainable. We could spell out even more starkly what ignoring that situation means, if you continue to ignore it for election after election, for children and grandchildren, people who are alive today, and what they will face a few decades down the road. That would bring it out more starkly. This is not just a chart that looks all right until two elections from now and we will worry about the future after that. It might be on the OBR to be a bit starker on the reality of what it would mean to follow these unsustainable paths for a few decades. It is not good.

We were all struck when Laura van Geest talked about the pre-measures forecast and how they had threatened to publish it if it was leaked, but now that has become the norm. What should happen? Would it be a good change to publish that ahead of the Budget?

The publication of the forecast, currently in legislation, is set by the Chancellor. It is a decision for the Chancellor, not for us.

We are a Committee that has to make recommendations.

There is a trade-off between transparency and unhelpful speculation that it could set off. One point here is that we have a fairly unusual Budget calendar in the UK. You have the spring forecast usually in March and then you have a very long gap until the autumn Budget, which is now the single policymaking event, in October or November. A lot of other countries do it in a different way. Instead of having a spring forecast, they have a summer forecast three or four months before the main Budget. That forecast is then seen more as a preliminary pre-measures forecast. It sets the ball rolling for the policy development process that the Treasury then undertakes ahead of the final Budget and forecast in the autumn. As I say, a lot of countries do that. You can see that it has advantages, potentially, in creating a stronger link between the first initial forecast and the final one compared to the current calendar in the UK where there is such a long gap between the two.

We have seen some chopping and changing. We used to have the autumn statement and the Budget. We now have the Budget and the spring statement. If we went back to the other way around, it would take a while to change the rhythm. Would that help to solve the problem?

You had a structure that was similar to this pre-2010. You had a pre-Budget report in the autumn and then the Budget in spring. The pre-Budget report was originally intended to be exactly that, a pre-Budget report or initial forecast to set the ball rolling for policy, and then you would have your final Budget in the spring. Over time, the pre-Budget report in essence became a Budget in its own right with a lot of policy around it.

There was a lot of discussion today with the other panel about resources for you. We have discussed this before. You are reliant on HMRC and the Treasury to quite a degree. Do you have enough resources to undertake proper analysis of tax policy or do you get what you need from HMRC, as it is?

We broadly do have the resources that we need to produce our core outputs. When our responsibilities have expanded over time since we were first set up, the Treasury has always given us more resource to do that. There are always areas where you would like to do a bit more, but, as I say, we can deliver with the resources that we have. We are smaller than a number of other fiscal councils, as people have told you, but that is because of the model where we use experts from HMRC and DWP. I think that works really well. Those analysts are deep experts on tax, benefits and the DWP benefits system. We have full access to their expertise and analysis. We have systems set out in our governance framework where we have the right to ask them to change their models or their analysis, if we judge that we disagree with it for whatever reason. We do that frequently, and we do not get any pushback on that. They will always do that if we ask them to. We have a very rich debate and a very rich set of analytical inputs from working with those colleagues.

You do not see any need for a change, then.

It is a very good model.

That is really helpful to hear. The mechanisms seem to be working for you. There have already been changes in this Parliament to the OBR’s remit from the earlier Act of Parliament back in July. Does the Chancellor’s fiscal lock strengthen your independence because any significant measures must be accompanied by you? Since that there has been some pushback against the OBR on how and when you score certain changes announced by the Chancellor.

It is certainly good practice that fiscal announcements are made alongside an assessment of the cost and a forecast to provide that transparency to Parliament and the public of the implications of government policy. We have seen the risks when that does not happen, and the fiscal lock reduces the chances that it will happen in the future. It has not yet been tested, but it is positive in that respect of ensuring—

In favour of being kept.

—that you have transparency around major fiscal announcements.

Well, we saw what happened in 2022 when that did not happen, which is why it was introduced. We also had some interesting evidence from the previous panel about costing opposition policies. The situation in the Netherlands is very different from that in the UK, as Laura van Geest put out, but just to be clear—and we briefly touched on this when we had you in front of us before—would you want the role of scoring manifesto promises? Perhaps manifesto promises are, as we heard from our previous witnesses, a bit too loose. Is there any scope for that? What resources would you need to do that? Is it something that you would ever want to do?

There is a case for better fiscal analysis of manifestos. There is a big question, as your experts have given you evidence on, as to whether the OBR is the right body to do that. It would draw us very much into the political sphere and create risks around our impartiality. There are also some quite significant, maybe not insurmountable but significant, delivery and operational issues. One is that, as Laura was saying, you would need many weeks, possibly even months, of interaction between political parties, where they gave the OBR their policy manifestos ahead of time, and they would have to be to a certain level of detail to allow us to cost them. It sounds like that happens in the Netherlands, but it would be quite a big change here. Secondly, going back to the point on our structure, we use government experts in HMRC and DWP to produce the costings. We do not do it in-house, so there would be questions around whether civil servants could do that for opposition parties under the current constitution. That is a question for the Cabinet Office rather than us, but, if we had to do that in-house and not use government officials, we would definitely need a huge number of people.

You would also have to have a threshold in terms of the size of a party. I typically have 12 people running in my seat.

May I just briefly add one thing on that? One of the big problems with it, which is not necessarily insurmountable, is just that manifesto proposals or promises are frequently just too vague to reach a fiscal judgment on. That is true even for bits of legislation that have gone through in the UK. The reason why the Employment Rights Bill has not yet been assessed or scored by the OBR is that it is still not clear on some of the most important parts of it. There are bits that are out for consultation, and they are probably the bits that, for either good or ill, have some of the most significant effects. We have not been in a position to judge something that has gone through Parliament.

That is very helpful. One other suggestion is that individual MPs, for example, could ask you to look at individual policies. It is a bit akin to the National Audit Office, where the Comptroller and Auditor General has complete independence. MPs and, indeed, anyone can ask the National Audit Office to look into something, he will then make the judgment about whether it is wise or we have resources to do it. He will make that judgment independently. Is that something that you would ever want to consider? Again, there might be a resource issue.

Again, it would need a change of legislation. Currently, our legislation says that we can look only at Government-announced policies. It would certainly require a big increase in our resource if we had to look at any policy that an MP put to us. Again, we currently do not do these policy costings in-house. They are done by Government Departments.

So the same problems apply.

The CBO, as your previous expert was explaining, is specifically set up to sit under Congress, as part of Congress, and has much greater resource to cost policies put forward by both parties in Congress.

It is a very different role.

It is a very different model from the OBR.

A big hot potato that has arisen over the last 12 months, with the spring forecast in particular, is who should be judging whether the Government have met their fiscal rules—the Treasury or the OBR. Have you any thoughts about whether that should change?

One gets back into the issue about bias and impartiality. The difficulty with the Treasury making judgments is that Treasury officials are civil servants. The question then is, “Who are they servants to?” It might be that they are servants to the Minister, and that puts them into a very difficult situation, having to say, “I am sorry, Minister. You have failed”.

So you are content to keep that role. That is fantastic.

It is the only way it works.

Yes, absolutely, just to make it clear and to get it on the record, because I know we have discussed some of these things before. We also discussed the size of the committee. We did not discuss it with our previous witnesses, but there are normally three of you. Is that the right size? What are the pros and cons of getting larger? The MPC, of course, publishes its own opinions, so you know exactly where everyone stands on something. Would you favour a bigger committee perhaps doing that, or is three the right number?

There are three of us, but we sit on top of an organisation with 50 people and we definitely encourage those people. There are some very skilful economists, analysts and people who know a lot about the fiscal situation, certainly more than I do. We certainly encourage an open debate about that. We do not need to encourage it because one of the reasons people quite like being at the OBR is that they are allowed, as you would not be in a more hierarchical institution, to come up to a member of the Budget Responsibility Committee and say, “You are completely wrong on that”. That is okay. It looks like there are three of us and we just sit in a corner and make a decision.

You are not God.

No. God does not live inside three portacabins on top of the Ministry of Justice.

The Holy Trinity.

They are not quite portacabins, but they are about the size of three portacabins. Anyway, you could definitely get them inside the entrance to the Bank of England, and have a lot of space left over. I will not bore you with all the people we talk to, but we spend so much of our time talking to the advisory panel. We meet representatives of the TUC, the CBI and other trade organisations. I spend quite a lot of time talking to investors in the gilt market and going to meetings in the City. We have regular meetings with think-tanks such as the Institute for Fiscal Studies, the Institute for Government, the National Institute and the Resolution Foundation. Right now, we are setting up some meetings, as we think about the long-term 50-year forecast, with a whole range of experts. Some of them are academics and some are not. We have just got the latest Office for National Statistics estimates of population and its new estimates of the level of net immigration to the UK. We are going to have a meeting with Alan Manning, who is probably the UK’s leading expert on immigration, and Madeleine Sumption, another expert, sometime very soon. It is in my calendar now. We really do spend a lot of time talking to a lot of people outside the OBR. If I could just come to the more specific issue about whether the BRC should be five people or seven people, the comparison with the Monetary Policy Committee is an interesting one. The Monetary Policy Committee is designed in a way not necessarily to come up with any consensus on policy. They all get one vote and the majority wins, and then people justify and show how they differed from the majority opinion. In the minutes of meetings, it now explicitly allows everybody to have their own paragraph. That could not really work very well for the OBR. If we had several views, and people set out their alternative scenarios for fiscal policy, there would always be one or two who are either much more pessimistic or much more optimistic than the rest, and you could guess which politicians would like which members of the extended Budget Responsibility Committee. At the moment, we reach a consensus, and I genuinely think it is not just a consensus between the three members of the Budget Responsibility Committee. It certainly includes pretty much everybody who works within the OBR.

It is interesting, because you could end up with a Supreme Court situation where members of the committee would be appointed for their political view.

That would be a danger.

Just as a quick point on this, it is the case that many fiscal bodies around the world have larger committees, but the two other fiscal bodies that produce the forecasts that a Government use—the Netherlands and Belgium—have small committees. The Netherlands has three people. Belgium has, basically, just a director who is an appointed leader, and then they work with the staff. As David has described, that reflects the nature of having to produce a very detailed forecast, which, as David says, would be very difficult to do in a large committee.

Just finally, is there anything else that we have not covered that you think would be a good change 15 years on as the new chair is about to come into post?

Both things have been mentioned. I very much agree that, if there could be more attention paid to the longer-run forecasts that we do, that would be a good thing.

We like the sustainability report.

That is partly up to us in the OBR, and we will do what we can on that. Also, I am not a great fan of traffic lights and all the rest of it. We do produce a traffic light-y thing. It is very pretty—and there it is. It is table 6.1 in the forecast we put out in March. You probably cannot see it, but there is some deep red, some light red, some amber, and bits of green. The problem with it is that, while it is a very good idea to take a more nuanced and broader picture of the fiscal outlook, if you just say, “Well, that is our result. Judge whether the Government are doing well”, some people will point to the green bits and some to the red bits. While it would be a very good thing to take a much broader view of the fiscal situation than just, “Where is headroom three or four years down the road?” I can also see the great advantage of having something that has precision and clarity.

I agree with that. Us trying to make more of the long-term projections and provide more information, in answer to the question earlier, on the wider fiscal position alongside our budget forecast as well, as we tried to do in the spring, and we can build on that in the future, are the key developments at the moment. We are also trying to do more of our forecast evaluation reports, which, as I said earlier, is a really important part of the transparency of what we do. Previously, we have done that on a year-by-year basis, but, now that we have a longer track record of forecasting, we can do deeper analysis, looking at a longer time period of our forecast record, allowing us hopefully to improve our forecasting techniques further. Those would be the main things.

Thank you. Can I thank both Tom Josephs and Professor David Miles very much indeed for their evidence? We will be seeing you again. In the meantime, we have Minister Torsten Bell coming in front of the Committee on 16 June for the final session of our inquiry into, “Whither the OBR after 15 years?” Of course, we are expecting the appointment of the chair of the OBR, and we will have a confirmatory hearing on that appointment when that arises. Can I thank, again, our witnesses?