Welcome to the Treasury Committee on Tuesday 15 July 2025. Today we have one of our regular sessions with the Office for Budget Responsibility, looking in particular at its fiscal risks and sustainability report. Then, we will move to the reappointment hearing for Richard Hughes, the chair of the Office for Budget Responsibility. He has been reappointed by the Chancellor, but this Committee ultimately holds the ring on that; we have a veto power over this appointment, so that will be a serious discussion on the basis of the questionnaire that Mr Hughes has submitted. I will ask Chris Coghlan to kick off the questions in our first session.

Good morning, Mr Hughes. Is there a debt level for the UK that we should worry about?

There is no hard and fast level of debt at which the Government should start to worry. In the recent past, you have seen Governments get themselves into trouble at lower levels of debt than the UK has at the moment. Ireland had a fiscal crisis back in the 2010s with a debt level lower than we have; Greece had one with a debt level a bit higher than ours in the noughties, and Japan has a debt level of 250% of GDP and is still managing to borrow on the markets, although it is under significant stress at the moment. The issues that you have to consider when thinking about safe levels of debt for any country include how exposed they are to shocks. Ireland was very exposed to shocks; it had a big financial sector, and when its financial sector went down, the country’s public finances were not enough to absorb that shock, and they had to go to outsiders for support. You also have to consider how quickly you can, and are willing to, adjust other policies to deal with stresses that might emerge on debt markets. In the case of Greece, they could not reduce spending and raise taxes to get ahead of the rise in their interest rates and the deterioration in sentiment. You have to think also about who is buying your debt and how secure the demand is for your debt. Japan has a very large and deep pool of domestic savings, which continues to buy Government debt, despite it reaching very high levels. When you look at the UK, in all three of those areas, there are reasons to worry. Our country is very exposed to shocks and has been particularly hard-hit by the last three shocks that the world has faced: the financial crisis, covid and the energy crisis. In each case, we took quite a hard hit economically and fiscally from those shocks. We have also raised taxes quite a bit in this country already; the tax burden is getting close to an all-time high, so we have already used up some of that policy flexibility. Also, the size of the state is getting close to a post-war high and it is already taking up a large share of public resources. As we highlighted in the report, there are concerns about the demand for UK Government debt. We have traditionally relied on the defined-benefit pensions market as a source of safe and secure demand for UK Government debt at reasonable prices. We think—we can go into it in this session—that that demand will wane in the coming years; it already has to some extent. That means that, to keep financing themselves, the Government have to lure in what we call more price-elastic investors to the market. That can have implications for the cost of debt and ultimately put some quite big pressures on the UK. There are reasons to be concerned about the level of UK Government debt and for it to feature in decisions that are being made.

Given that, how do you see the relationship between debt and growth?

The sustainability of debt depends in the long run on how the interest rate you are paying on your debt compares with the rate of growth of your economy. At the moment, we have one of the most adverse configurations of those two things. The split between the interest rate on our debt, which nowadays is getting up to 4.5% of GDP at the 10-year mark, which is where our average maturity is, is several percentage points above what the growth rate in our nominal economy is—well, about a percentage point above where our nominal economy is—so that just makes it harder to stabilise the debt-to-GDP ratio where it is.

Given that, what would you say would be the best use of spending to reduce debt?

As you know, we are prohibited from giving Governments advice about where they spend their money or how they raise their taxes to pay for it. In the course of our work analysing policy options, we do look at where we think areas of public spending can generate positive effects on economic growth, and areas where Governments have introduced policies that we think can make a meaningful difference to the drivers of growth and potential output. Measures that boost the participation rate of the population boost employment, which boosts growth and then boosts tax revenues. Also, measures that Governments have taken to boost business investment have also built up the capital stock and increased capital per worker, which we would expect to have some impact on productivity. Finally, and most recently, in our last EFO we looked at reforms to the planning system, and the EFO before that looked at reforms to public investment, both of which we think can also make some ultimately modest, but, if sustained, significant differences to the growth rate of the economy.

Given all of that—given that growth could reduce debt—does it concern you that the Government are only increasing R&D by 0.7% a year over the next five years?

These are issues for Government. They always have to make choices and trade-offs about where they spend their money. R&D drives potential output by increasing the stock of intangible capital, so it is one of the things that we take account of when thinking about the growth potential of the economy, but as I said, the Government can pull a number of different levers. They also need to bear in mind that higher and higher levels of taxes are also not good for growth. If higher research and development is being funded by a higher tax burden, there are some choices and trade-offs that need to be paid for.

Unless it is funded by debt. Finally from me, do you think it is responsible of the OBR to have a productivity forecast that is two thirds higher than the Bank of England’s, given that that implies an extra £50 billion fiscal black hole for the Chancellor if you are wrong?

Our productivity forecast is something that we keep under review. We are in the midst of what we call our supply stocktake, where we look at the drivers of the potential output side of the forecast. In the near term, we are reasonably aligned with the Bank of England, but, in the medium term, there are some differences between us and the Bank—and some other forecasters. That is an issue that we keep under review. When we look back at our forecasting record over the last 15 years, in general, our GDP forecasts have been pretty much in line with the consensus, although both turned out to be wrong—both us and the consensus—because it turned out that we were both too optimistic about the outlook for growth. We come to our own judgments on these kinds of things. As I said, we have a review under way. When our forecasts entirely align with the Bank of England, people tend to complain that we are all subject to groupthink; when they differ from the Bank of England, people tend to ask, “Why do you guys differ on the same question?” From time to time, we and the Bank diverge, and we currently do in the medium term, but it is an issue that we keep under review.

Mr Hughes, in terms of fiscal risk versus debt burden, if we underestimate the growth impact of public investment, it damages our debt burden, right? If you at the OBR underestimate the benefit of public investment, it means less investment and less growth, and then, fundamentally, that ends up being the doom loop. Do you believe that the OBR currently adequately accounts for the growth impact of public investment?

I think we do. We did a pretty comprehensive survey of the literature on this question. We also looked at what other, longer-established forecasters apply in terms of elasticities between levels of public investment and levels of capital stock and then economic growth in the long run, and the sorts of ratios that we have applied to the contribution to public investment to growth match quite closely both what the literature suggests and what our sister institutions in places such as the US, the CBO, apply to measures pursued by the federal Government. As always, we keep those estimates under review and look at the impact that they have. It is important not to overstate what the Government are doing on public investment. They are, in essence, keeping it constant as a share of GDP. The previous Government planned to reduce it somewhat over the next five years, but this Government plan to keep it more or less constant as a share of GDP. That would continue to support the level of economic activity that we have. There is not some take-off in public investment over the next five years that will somehow support additional levels of economic activity; they are just not cutting it back, as a share of GDP. We did think that there was some growth benefit even from maintaining current levels of public investment, but you have to remember that the Government have invested about 2% to 3% of GDP per year over the last 15 to 20 years. The stock of capital supporting the economy is about 200% of GDP, so it takes quite a long time for the marginal addition to the capital stock made by the Government to make a material difference to the growth prospects of the country. But we try to show in our forecast that if over a five-year period it makes a modest difference of tenths of a percentage point of GDP, over a longer period—if we go out 10, 15 or 30 years—we can see a bigger difference because that capital stock builds up over time.

On the other side from public investment, there is public spending. To be fair, I was not entirely clear in my question. Professor Miles, I want to talk about a specific action of public investment, if you like, or public spending. The last time you came before this Committee I asked you, I think, three times why private building under the planning Bill would lead to higher growth, but public investment in public building and social housing would not. Page 27 of the EFO states that the expected increase in productivity “is driven by the residential planning reforms, which we expect to increase construction sector productivity and housing services due to the higher housing stock.” The EFO is quite clear that the OBR’s view is that private sector building, induced by the planning reforms, leads to more growth in the short term. But at the moment, in the OBR forecast, the same is not true of social house building. Last time I asked this question three times; I looked back at the transcript, and I do not think I got an answer. Do you have an answer now?

I think that all house building, whether or not it is funded by the public sector—local authorities or the Government—has the potential to increase the level of GDP. In a way, it happens almost automatically, because if you increase the stock of housing, in GDP you count either the explicit rent paid, if it is rental housing, or the implicit rent on properties that are owner-occupied. That is part of the measure of GDP. There will be an impact there, so it is not the case that housing that might be in the rental sector, and might be funded by local authorities or by Government, does not have an impact on GDP.

I am asking specifically about your forecast. In that forecast you have an increase in potential output induced by the planning reforms leading to more private building. I am saying that, at the moment, in the EFO, your forecast says that if you build more in the private sector, you increase trend productivity growth—you have more space, if you like, both economically and fiscally—but, in the forecast, the same is not true of social house building.

Part of the impact on productivity is what I just described: if you increase the housing stock, just the number of houses, that will show up in GDP. That is in our forecast. There is also a different effect from the planning restrictions changes allowing some land to be built on that, at the moment, cannot be built on for residential purposes. There is a productivity impact from simply taking land that at the moment, let us say, is agricultural land. At present, the value of that land and the contribution it makes to GDP comes just from using it for agricultural purposes. It might cost £7,000 or £8,000 an acre. That can turn into £300,000, £400,000, £500,000 or even £1 million an acre if you allow residential development. In our forecast, that is the bigger boost to productivity that comes from house building. It would not matter whether the houses built were public or private sector housing. You would still get the boost from taking land that was relatively low productivity—it has sheep on it—to housing, the value of which is clearly much higher. The income thrown off by the housing—in either actual or implicit rents—is just much higher. That is the mechanism in our models.

Last time, I think we came to this point: at the moment, 1 million houses have planning permission but are not being built. I said to you then, if the public sector was to build those houses, at the moment the OBR forecast model does not say that that would improve growth in the short term—it does not lead to fiscal space. That is the point about the doom loop. In one sense, you are undercounting the impact of public investment.

I am not quite following the question at this point.

Let us say that I have a piece of land; this land that is currently undeveloped but has planning permission. I build social housing on it. At the moment, in the OBR forecast, if I build social housing on it—

I don’t think that is right.

It doesn’t lead to more growth?

I don’t think that is correct; I don’t think that would be true in the way we model things.

So you are saying that the increase does not come from construction. Either it comes from construction or it doesn’t.

When you construct more houses, they will generate higher measured GDP in the future, for the reasons that I explained—because they will throw off either explicit rent or implicit rent if they are owner-occupied. That is part of GDP.

John Grady is going to ask about pensions later, but I want to talk about the impact of some of the changes you point to: the transition from DB to DC pensions and the implied changes in the profile of gilts. I think the OBR says that you push up interest rates on Government debt by around 0.8% if the debt level remains around 100% of GDP. Is there any way of avoiding that?

It is difficult to avoid the slow sell-off, which will take many decades, from DB pension funds and their ownership of gilts, because DB pension schemes in the private sector are overwhelmingly closed to new inflows and they are going to run off over time. At the moment, those DB pension schemes have quite large holdings of gilts, and it is almost inevitable that that will run off gradually over time. It may be replaced as the size of defined-contribution pension funds grows to some extent to offset the decline in defined benefit, although defined-contribution pension schemes have in the past held and continue to hold very many fewer gilts, as a percentage of their portfolio, than DB schemes.

Given that they are not obliged to make a fixed payment in the same way as DB schemes typically are, it is not inevitable that the trajectory would work like that, is it?

Well, they would become something different from DC schemes if that were true, since in a sense the defining characteristics of DC schemes are that the money is paid in, it is invested in assets and those assets are the source of the revenues for the pensions. If the assets do not do very well, the pensions paid out will be lower, and if they do really well, they will be much higher. So they are not fixed. That is part of the reason why DB schemes were natural buyers of gilts—because they wanted to try to hedge a fixed obligation to pay pensions against the assets. DC schemes are not in the same position.

One of the commentaries I have seen recently suggests that the way the Debt Management Office operates, with a greater emphasis on long-term linkers, has been expensive for the way our debt is paid for. I recall as a Minister—I think my colleague Dame Harriett would have done the same—engaging with the market to assess the appetite, and then the DMO essentially does rounds of gilt issuance to reflect that demand pattern. Can you say how we could do things differently, or should have done things differently, to avoid the situation that we are in, with a disproportionate number of long-term linkers in our profile of debt?

I would say that since index-linked bonds were first issued, which was probably almost 45 years ago, in the early 1980s, they have on the whole generated reasonably good value for taxpayers, because the real interest rate on the index-linked bonds, although high when they were first issued, came down very substantially over that period. Until relatively recently, they had a real interest rate of minus 2% or minus 3%, so the UK Government—not today but a few years ago—could have locked in 30 or 40-year borrowing at a real interest rate, an inflation-adjusted interest rate, that was negative. That is not quite the same today. Those yields have gone back up; they are positive again, and it is not so obvious that they are such great value today. I think the reason why the DMO has skewed its issuance slightly away from index-linked bonds recently comes back to the defined-benefit pension schemes, which have historically been the big buyers. Their demand is already beginning to drop off to some extent. I suspect that this will be a trend that continues into the future. For good or ill, I think the Debt Management Office, in trying to match what it issues against where the demand is, is very likely to be issuing a lot fewer index-linked bonds in the future.

The Chancellor’s new debt measure generates £113 billion, which is committed to extra capital expenditure. How do we measure, and how do you account for, the return on investment? Where is that reflected, so that the taxpayer watching the Committee today can see it? Where can the ROI be seen in the public accounts?

In two places, I guess. One is the direct financial return that the Exchequer would get from investing in financial instruments. To the extent that they are lending money or investing in equities, the loans would generate interest as well as repayment of the principal, and then the equities may well appreciate in value and may generate dividends. The financial assets that are held by things like the National Wealth Fund will generate some kind of direct investment return for the taxpayer. Then there are the potentially indirect benefits from a faster-growing economy that you might get if moneys are invested in schemes that boost the growth potential of the economy and boost tax receipts.

But it is not really explicit, is it? The narrative of the Government will be, “We are investing extra money to fix the mess that was left and invest in productive capacity,” but in terms of a consolidated, neat view of the net impact of that investment, it is quite piecemeal and difficult to put together in one place and be assured of what the return is. Is that not a reasonable assertion to make?

That is right. Oftentimes, the term “investment” gets used to cover an awful lot of different things that the Government do.

Yes, some of it isn’t—under my Government too, I am sure.

Sometimes they are talking about direct financial investment. Sometimes they are talking about building a physical asset, like a house or a railway, some of which make money and some of which lose money, depending on the way you regulate the prices and the rents. Sometimes they just talk about investment as a general term for spending money on public services, which can include education, health and everything else. Some of those investments, like financial investments, do generate very direct financial returns to the Exchequer, and those can be measured based on what they charge in interest and how they perform as instruments. Others—

I am just wondering if you think there is a better way, given the policy shift we have seen, of giving the public assurance over the effect of that, such that they can say, “Well, we voted in this Government. They’ve changed these rules—they were perfectly within their rights to do so—but what’s consequential of all this extra spending/investment?” Is there a better way that we could do it, and should you look to evolve a more coherent measure so that the public can have more confidence?

We have tried to evolve as quickly as we can. In terms of capturing those direct financial returns, we have a much beefed-up chapter on the balance sheet in the fiscal risks and sustainability report, which shows how those investments’ returns, but also potentially the risks—the Government are also taking some risky investments and making some risky loans and they could lose money, so it is not all about upside, necessarily—can go. In terms of providing much more transparency about the Government’s financial balance sheet and those direct returns, we are providing a lot more information and are happy to provide even more if you are interested. We have got some people who are very enthusiastic on these issues in our organisation. Also, with those indirect returns, which were previously implicit in our economic forecasts, we have tried to make it much more transparent and explicit about the way we measure the economic effects of Government policy, be it things like the expansion of childcare and what that does for employment, or public investment and what that does to boost the capital stock. We do try to answer those questions to the extent that our models and the empirical literature can give you reasonable answers, but if there is more that we can do, we are always in the market for interesting ideas.

In some senses this is going back to the first question, but I want to talk about why the price of UK Government debt is so high. It feels like at least one of the reasons is that the markets are starting to worry about the fact that every Government says, “This is what we’re going to do over five years, and all the good stuff that you’re going to be happy about, like getting spending under control, happens in years 3 and 5,” but it happens on a rolling basis, so the market’s faith in that strategy ever paying off deteriorates. Chris asked whether there is such a thing as the right debt level, but should we be doing something about the fact that these are rolling forecasts and taking account of the cumulative effect of that?

One of the things that we highlighted in the report, and that you might have noticed in one of the charts, which shows basically what was forecast—

Is this the yellow and green lines?

It is the yellow and green lines. It was Ed Conway’s favourite on Sky News. It basically shows what the Government claimed was their plan, which shows debt rising for two years and then typically falling for the remaining three, and then what turned out to happen in practice, in out-turn, which is that the debt rose and rose. Part of that was the reversal of planned policy consolidation, be that the reversal of planned tax rises or the reversal of planned spending cuts. Another part of it was that another shock came along. You do have to have some sympathy for policymakers; they have been making fiscal policy in a very volatile environment, so some of that higher debt level is that covid and the energy shock came along. Another part of it is that, since the economy and the public finances stabilised, there has not been a determined effort to reduce the level of debt over time. There has been, in practice and de facto, a willingness to allow its level to get ratcheted up and up over time and then to commit to getting it to fall on a future date.

Are our inflation expectations a contributing factor? When we had the spending review hearing, we learned that the forecast of real-terms increases in spending are based on inflation remaining at 2%. It frequently seems to be going higher than that, even though we do not expect it to, and Governments do not really want to cut spending in real terms, so they are going to keep chasing that inflation figure. Is that 2% an inbuilt assumption? Is that incorrect, or do we always have to target that because that is what the Government target?

It certainly matters for things that are indexed to inflation, such as welfare benefits or the state pension. As I am sure we will come to, the state pension triple lock has turned out to be a lot more expensive than we thought, because inflation has surprised on the upside. Inflation tends to surprise much more on the upside than it ever does on the downside, and the triple lock does not allow you to take the benefit of it surprising on the downside, because of the 2.5% floor. For other welfare benefits, it does not surprise on the downside to the tune of double digits the way it has surprised on the upside more recently, so that becomes a ratchet for welfare costs. For departmental expenditure limits, it is a process of the Government and the Treasury basically trying to hold the line on a set of cash budgets while inflationary pressures build up. We have seen in recent years that you can hold that line for a little while, and then it tends to pop out the top, in the form of either industrial action or continued difficult pay negotiations for Government.

I thought it would be worth picking up on the rest of the chapter on debt. Paragraph 2.75 of your report notes that the switch from DB to DC is likely to lead to a permanent drop in domestic purchases of gilts from around 30% of GDP to 15% by 2070. You go on to say that, even at today’s debt levels, you are looking at an additional interest bill of £22 billion a year. I have two questions. First, can you talk us through how much of a problem this becomes if debt goes on a higher trajectory, as your reports suggests it will? Secondly, what policy options do the Government have to address this?

The calculation that we did, as you rightly say, assumed that the amount of debt relative to GDP stays the same but there is less demand from the pension sector. Then, you would have to induce other people, who at the moment are not holding those bonds, to hold them. That almost certainly means you have to offer them a better deal—a higher interest rate. That is where that calculation comes from. Of course, that would all get somewhat more tricky if the stock of debt was not staying at 100% and were to be rising. The estimate is about 0.8% on the interest rate on Government bonds over the long run—it is a slow process—but, in an environment where the debt were rising on top of that, it would add to that number. That is, in a sense, one of the reasons for our projection that, given the demands on Government spending and the current system of taxation, it looks like there will be, if nothing else changes, a rising level of Government spending relative to GDP, and a constant tax take at what is already quite a high level. That obviously then generates deficits, which add to the stock of debt, hence the simulation that says that if nothing changed, it would not be right to assume that debt just stayed at 100% of GDP; it would indeed be rising over time, ultimately probably in a way that would be unsustainable, because there would be a feedback on the interest rates.

If I have correctly understood the report, in essence what you are saying is, “If public spending and tax continue on the same trajectory, debt goes up significantly, but, look, you’ve got another problem: you’ve lost a big buyer for Government debt.” That is really what you are saying, isn’t it? It becomes unsustainable because you have also lost a buyer as well, which causes a real issue.

It is right that there are two separate issues, both of which, as one looks far into the future, are worrisome. One is exactly what you say: one of the strong buyers of UK Government debt over the past several decades is going to be declining. Then there is a separate issue on top, which is the pressure on spending against tax revenues.

Thank you, Professor Miles. Mr Josephs, you have a long-standing interest in this issue. Is there anything you would like to add on the topic of losing the DB buyer of gilts and what that means?

One part of the analysis that we do in the report is to consider what that implies in terms of the type of buyers that might have to step in to replace the demand from pension funds. We do some analysis there, based on some studies on the price elasticity of different types of buyers of gilts. It is basically the case that the defined-benefit pension funds have had a structural demand for gilts because they match their liabilities so well, whereas the other buyers that you might have to attract into the market could be, for example, overseas buyers, who have a much wider range of assets that they can invest in and therefore are likely to demand a higher price. That is what leads to the estimate in the report of the increase in gilt yields that might result from this over time—it is a long-term process—of around 0.8 basis points. We do highlight in the report that there is a lot of uncertainty around that estimate. It depends on the types of buyers, and on the degree to which it has already been anticipated by the market and so is already in prices, which may well be the case to some degree.

Presumably 0.8 is at 100% of GDP.

Yes.

At 270% of GDP, 0.8 is probably a little low, isn’t it?

Yes, exactly. If debt were rising and you needed to attract in even more buyers, it is likely that there would potentially be a bigger effect.

As to what has been driving the expansion on the liability side, we have discussed that an awful lot. It is basically that the Government have been issuing lots of debt, mostly in the form of gilts, although it is probably worth also highlighting that one of the phenomena since 2008 has been that some of the gross issuance of gilt by the DMO has been bought up by the Bank of England through its quantitative easing operations and replaced as a liability by the central bank’s own liabilities in the form of central bank balances. On the liability side, things have gotten more complicated. We are used to having boring, sometimes index-linked but generally speaking long-dated liabilities where the interest rate does not change very quickly over time because you are not replacing that much of it that quickly. When the Bank of England came in with QE, it basically replaced a liability with a very long maturity—average 15 or 16 years—with an overnight liability, because commercial bank deposits to the Bank of England just pay Bank rate. That has already made the liability side of the public sector balance sheet a lot more sensitive to day-to-day changes in interest rates. You can see that in our fiscal numbers. Our interest costs have gone up a lot. That is, first of all, Bank rate going up, and then, over time, the overall liabilities of the gilt portfolio catching up with the market rate for gilts. So the liability side of the balance sheet has gotten more volatile and more sensitive to interest rate changes. The asset side has also become more complicated and certainly more risky. The UK Government has not typically held very many financial assets. It acquired a bunch during the financial crisis because it bought a bunch of banks, but over time it has disposed of those. It has also accumulated a large stock of student loans, which have a variable rate of repayment, because it depends on the future earnings of students and several other parameters. So the student loan portfolio also carries some risk, because you do not know how much of it is ultimately going to be repaid; it depends on what students earn. Over time, the Government has also become an active lender in a number of areas. It lends money to businesses, it lends money for energy projects, and those sorts of things. Those loans carry risks because businesses can default on them, and if the Government do not get repaid, that also has consequences. We have tried to illustrate that the Government is now targeting the financial balance sheet as one of its fiscal rules; there are many more sources of things that can go wrong in that than just looking at the debt stock.

You are right, it is complicated, because you then have to value the assets side and the liabilities side of the public sector in ways where some of the assets are quite volatile and difficult to judge. It is difficult to judge things where there is not a liquid market. You do not quite know what proportion of loans used to finance investment are going to be repaid in full. Some of the things are easier to value. For example, when we move to public sector net financial liabilities, you bring into the assets and the liabilities side some defined-benefit pension liabilities of local authorities, which is a big number, and their assets. Some of those assets are very easy to value, but they are very volatile. I think there is something like £500 billion of assets brought in, a high percentage of which are equities. That gets very volatile and one needs to assess the riskiness of that as you project where overall net liabilities of the public sector might go. On the modelling side of the impact of greater investment that might be encouraged by off-balance sheet financing, if there is more physical investment in infrastructure, for example, we treat that in the same way as any physical investment in infrastructure. It has the potential to boost growth in the economy and the productive potential of the UK. The balance sheet challenge for us is that some of the assets are more difficult to value now than when, in a sense, the public sector fiscal target was defined in terms of things that were easier to value. That does not mean that is the wrong thing to do. It just makes it more challenging to model it.

I would say that the public sector net financial liabilities, which is a wider definition and takes more things in on the balance sheet than the asset and the liability side, is more difficult to finance and will be somewhat more volatile. That does not mean it is not the right thing to do. It just makes it a bit more difficult and quite how it will evolve is a bit more uncertain and more subject to shocks. For example, a significant sell-off in the equity market could move the net financial position of local authority defined-benefit pension schemes by £20 billion or £30 billion very quickly.

People seem to have lots of expectations for our forecasts. We have taken on two new people to work on these issues. That is, for us, a significant increase in resources to look at the assets and liabilities side of the balance sheet. We do work quite closely with the people who oversee the local government pension scheme, the National Wealth Fund and the British Business Bank, who also have to do those kinds of projections for their own financial accounts. The Treasury has put in place a framework for monitoring and evaluating these things. We work quite closely with the rest of Government to try to put these forecasts together, but it is a challenge—it involves judgments and it creates additional risks. One of the reasons why we did a whole chapter on it in this report is just to highlight that those things need to be taken into account and thought about.

I want to go a bit deeper into the report and ask about the British Business Bank specifically. You have mentioned in there that the British Business Bank currently expects to lose between 30% and 40% of its loans, and that you are going to be updating the way that you use the methodology to reflect those kinds of losses. What sort of impact is that likely to have on the outlook?

There are a number of different ways in which you can treat loans and in the ways in which these institutions account and we forecast. Where there are portfolios of loans where there looks to be an expectation of significant losses, we do expect for the ONS to impose some write-down on them, which is the treatment that they have used for student loans where there is just a significant expectation of not full repayment. For some of the British Business Bank’s loan portfolio, especially that geared towards start-up loans in more speculative enterprises, you would expect to have a pretty significant default rate, so the ONS is looking at what the right rate to apply to that group of loans would be. We are discussing that with them, and we will reflect that in our forecast, as well as what we think the dynamics are likely to be going forward. At the moment, that portfolio of loans is only a few hundred million, which is a lot of money, but in macro and fiscal terms where things are denominated in billions and trillions, a write-down on those loans would make a bit of a difference to the outlook for PSNFL, but it is nowhere near making the difference between meeting or missing the Government’s fiscal rules. The other thing to bear in mind, to come back to Mr Dean’s point, is that PSNFL is about it falling in the medium term, not about the level, so accounting changes that move the level up and down matter less for whether the Government are meeting their rules than for whether they affect the trajectory of it over a five-year period. If over a five-year period, you expect a constant default rate on the loan portfolio, that does not really affect the overall trajectory of where the balance sheet is going. It is only if you expect an increasing default rate over time or something causes you to change your view of the outlook for these kinds of instruments, which gets worse over time, would that make a big difference to how the Government have expressed their rule for the PSNFL.

On the use of including the losses, do you think it is the right time to be bringing that in, given how much data there was on the student loans losses before that was included? Is it appropriate to bring this in now when the Government have just invested so much money in the British Business Bank to try and incentivise it? Could including this kind of speculative loss have any negative effects?

My general philosophy is to recognise information as soon as you have it, and to not live in denial of these sorts of facts. Particularly with a portfolio of loans that are designed for start-up enterprises—where the business itself is expecting a significant default rate and is factoring that into its own accounts, its financial decision making and risk taking—we know that forecasters should factor that into their forecasts. Otherwise, you are planning on inconsistent bases. It can create a real risk, which we have seen in the UK and other countries, where if you have a perverse accounting treatment of something, it encourages the use of that instrument, because of the way it is treated. It can drag resources in favour of loans because people think, “Well, loans always get treated as though they have full repayment, even if they do not.” You would not want those kinds of incentives to be created within either your accounting or your forecasting system.

It just seems strange that speculative losses are included but speculative gains in some other areas might not be scored, because you have not seen the evidence. It would be interesting to see whether 30% or 40% is the right amount, and then we can have that conversation in a year’s time.

We have some experience from covid loans, and you saw pretty high default rates on those. It is important to bear in mind that it is hard to make money on a loan, because people are not going to pay you back more than they borrowed. There can be speculative gains on equities and other kinds of investment, but for these sorts of things, 100% is usually the best that you can hope for.

Mr Hughes, I note that in your fiscal risks report that you have gone into a bit more detail than usual on the risk of water companies being potentially reclassified as public sector. As a south-eastern MP, my constituents suffer the financial risks of Thames Water every month, so it is a very live issue.

It is for a lot of us.

It is for many of us on this Committee. Do you go into more detail because you are starting to hear more from the ONS on the risks of your classification? Could you tell me a bit more about how you see that changing over the next year or so?

I guess there are a few things. We do not highlight just water; we also look at housing associations and higher education institutions. Especially for things like water, as they are providing an essential service, it is a very heavily regulated sector. It is one where the Government are imposing more and more regulations on corporate decision making. We have seen in the past that, at some point, the level of Government intrusion into corporate decision making in a sector gives the Government effective control, and then that gets reclassified as public sector. That is what happened with Network Rail, and that is what happened with housing associations, until the Government took some controls off to get them back into the private sector. Often times, that imposition of controls is geared towards managing financial challenges in the sector, because the Government are putting more money in and they want assurances about getting repaid. At the moment, we know that the ONS is looking at the classification of the water sector in general, as well as the Government taking additional interest in some water companies, and there are some concerns that some of them might even end up in receivership. There are a number of reasons to believe that at some point, either because of a methodological decision of the ONS or because of the financial condition of some water companies, they could end up being taken into public administration. The point that we wanted to make by illustrating that example here is that, once that happens, you have to pass that balance sheet up into different bits, if you want to see what it is going to do to the Government’s financial balance sheet. Water companies have a lot of debt, and they do not have that many financial assets, so their financial balance sheet impact is quite adverse. However, they have a big network of pipes and reservoirs, which does not count towards PSNFL, so it can be quite a negative financial balance sheet hit, despite the fact that, overall, the assets and liabilities of water companies are more or less in balance. It is just that, from a PSNFL point of view, their assets are the wrong kind of assets—they are fixed assets not financial ones.

On that point, you wrote in the 2019 Resolution Foundation paper about the public sector net worth as a target. As you describe in the fiscal risks report, there are significant financial liabilities that are captured in PSNFL, but there are also non-financial assets like the pipes and the physical infrastructure that are not, but they would be captured in the broader net worth measure. Looking at all of this, how much does the reclassification of water companies really matter when it comes to things like the debt rule and a broader valuation of what the Government are looking after?

I should say that I have no official position on what fiscal targets the Government should have, regardless of what I may have written in the past. [Laughter.]

The Torsten answer!

Our job as people who run the OBR is to assess whatever fiscal rules the Government have and highlight their features and discuss their risks. What we also try to do is monitor a broad array of indicators of the financial health of the Government. We have not taken our eye off public sector net debt and the targets that previous Governments have used, because we think that really matters. It is really important what you are on the hook to repay your creditors for. This Government are also interested in what is happening on the financial side of their balance sheet and want to make more active use of that. We are happy to look at that. We have the wherewithal to forecast it. But we also try to highlight the features and risks of that as a fiscal target. David talked through a number of them around valuations. We also forecast public sector net worth. You can have a look at what that does over time. What is interesting about all of them is that they all more or less do the same thing: debt has gone up, financial liabilities have gone up and net worth has gone down, all pretty much in the same direction and all for pretty much the same reason, which is that the Government’s been running a big current deficit and they have been funding it by issuing gilts. Everything else has bumped along in the background.

Are you currently making preparations to forecast the financial liabilities of water companies to get ready for potential reclassification? Could you talk us through what the different steps would be in the next few years if something like that did happen?

We try to be ready for anything. I am certainly not signalling that this is imminent or that it will happen, ever. But what we try and do is keep an eye on where the risks are around the Government’s perimeter, because we know that the ONS can make sudden and dramatic decisions. If it is coming up to a time when we have to do a forecast, we want to be able to consolidate those with our forecast and easily reflect them in the way in which we do our financial projections. We are always surveilling the boundary of the state, as it were, to look at where the risks are. We did that a lot with Network Rail back in the day. We did it a lot with housing associations. With this chapter we also wanted to highlight who is camped out on that frontier at the moment and what the risks would be if some or all of them were to find their way into the public sector.

Do you have a view on how PSNFL would change over the next few years if any water companies were reclassified?

My sense is that it would be a level effect primarily, rather than one that would materially affect the trajectory of PSNFL. Because the target is about the trajectory in year 5 at the moment, stuff that moves the level up and down is less exciting. But that does depend a bit on the structure of the water company’s debt and whether they are paying a much higher interest rate than the rest of the Government. That is not the case at the moment, but it depends on what terms they start to borrow.

You mentioned the significant impact it would have on PSNFL, but it would have a lesser impact if you moved to a different set of fiscal rules than public sector net worth. If you were to bring the water companies on to the balance sheet, you would say that the obvious thing to do is to switch to that role. But what would be the advantages and disadvantages of doing that?

As I said, we forecast whatever rule the Government have. I am not sure it would be an obvious thing to do to suddenly start focusing on net worth just because the water companies have come in. Water companies are important, but—

Some might argue that that is the way to solve the problem.

In the end, they add a few tens of billions one way or the other to a balance sheet that has trillions on either side. There are pros and cons to any fiscal measure. On the pros of net worth, it is a more comprehensive picture. The cons are that quite a lot of fixed assets of Government do not yield any kind of serious financial return to Government. To think of them as something that can somehow pay for the debt that is issued to finance them might not be the right way of thinking about things. Government assets are not like corporate assets. They are not all there delivering earnings for the business in future. A lot of them are just delivering wellbeing for citizens, but they are important and they are good. The Government spend a lot of money on military equipment; that is important to keeping our country the way we like it with the borders that it has, but that is not somehow generating a financial return that will pay you back for having all that military equipment at your disposal.

On that, Mr Hughes, a key part of the Government strategy is around defence R&D, which economic forecasts show does increase economic growth. In fact, it is one of the most significant ways of increasing economic growth. Would you agree that if they are spending it on defence R&D, it is a good use of Government money for boosting growth?

To the extent that it feeds into the Government’s contribution to the capital stock as an intangible asset, it would be one of the things that goes into our calculation of the drivers of potential output—I would put it that way.

The fiscal risk registers in your reports paint quite a gloomy picture, and persistently so. Last week you said that there is a trend of “debt ratcheting ever upward” and that you see debt as a percentage of GDP going up a further 5% over the next five years. In the absence of improved economic growth, it is difficult to see any change in that trajectory. Is that the right conclusion to draw?

Governments have policy choices as well—they do not just have to hope that the economy does better. They can make sure that they have the debt level they feel is secure and prepares them for future shocks. It has certainly been the case that since we started producing these reports in 2017, a lot has happened, and most of it has been bad. Covid delivered a big shock to the Government—that raised debt. The war in Ukraine further added to our stock of debt. We have seen disappointing economic growth. More recently, we have seen a very dramatic and rapid rise in interest rates on that much higher stock of debt. All that has put the public finances under a lot of stress, so you can understand why Governments have found it difficult to deliver the kind of fiscal consolidation that gets debt falling again and rebuilds the fiscal space. The concerning thing if you look at report after report that we have produced is that the number of risks has mounted over time. War has started in Europe again. We have a US Administration that is putting tariffs on everybody, although we have managed to carve out a few exemptions for ourselves. Global risks are rising and our capacity to absorb them has been shrinking because we have ended up having to spend an awful lot to deal with those risks.

Let us look at what Governments can do. I think we all agree with your assessment that what has happened over the last seven, eight or 10 years is quite significant. At the spring statement, you were criticised for not having performed the productivity impact assessment of cutting welfare spending in time to put it in the forecast. That proved to be a pretty good judgment, given Parliament’s decisions last week. However, there is concern over exactly how you make the judgment—my colleague Lola McEvoy touched on this—around the arbitration point on what you include and what you do not include. Some pieces of legislation, such as the Employment Rights Act, will have an impact assessment of nearly £5 billion. There is a range within that. Different pieces of legislation will be at different stages across the Commons and Lords. Can you explicitly nail for us who makes that judgment? Is it the Treasury? Is it you? What is the cutting-off point? You said that you received data very late. At the Liaison Committee, the Prime Minister was critical that it was not scored. Is there anything you would like to say about how that process works across Whitehall?

I will make some general points about how we deal with what we call supply-side effects of policies, which can be their employment, investment or growth impacts. Mr Josephs might want to say a bit about the specifics of the welfare case. What we need first from the Government is a policy. If the Government do not have a policy, they can explain to us, but we cannot put that in our forecasts. In the case of the welfare reforms, the reforms to benefits came to us very late and past the deadlines that the Treasury had set. There were no specifics on the employment support programme. They could not tell us which groups were going to benefit from that programme or what kind of support they were going to get. There was no policy for us to score at our forecast. In terms of taking into account the employment effects of that policy, the detail was not there. Once we have the details, we have to make a further set of judgments. We apply a set of criteria, which we set out in a paper we published two years ago. Will the policy have a supply-side effect? Is it a significant policy? Will it have an effect beyond the ordinary day-to-day business of Government? Will it have a durable effect—is it not just a one-off policy, but something that will last for years and support potential output in future? Most important and, in some cases, most challenging, is whether there is evidence to support what the Government claim will be the long-term economic effect of the policy. Where those criteria have all been satisfied, there is a clearly defined policy and those other things. We have, from time to time, reflected those on the supply side of our forecast, because people have asked us to. People have said, “You guys cannot be growth nihilists. What kind of Government policies can have an impact on growth?” Both under previous Governments and under this Government, we have done that. In the case of the welfare issue, our main issue on the employment support policy was that the policy was not defined. So the rest of those questions did not enter into it, because you couldn’t answer them. In terms of the specific timing, Tom might want to say more—

Can we just go back to the welfare question? In your report last week, you say that if the current gateway of onboarding people continues, there will be £12 billion extra on welfare by 2029-30 compared with the current time. Do you have quantified options of what choices Parliament and Government could make that would address that? Part of the problem here is that you are making a commentary on things as is. You studiously—understandably, because that is what you are legislated to do—refrain from making policy judgments. But perhaps it would be helpful for those watching to understand what choices could have been made that would ease that pressure and stop it going up even further than you predicted in the previous report.

We are definitely legally prohibited from answering the second part of that question. We are prohibited from providing alternative options to Government policy. We can only score the impact of Government policy. But Tom and colleagues at the OBR have done lots of work on what has been driving those trends in rising spending and claimant numbers in disability benefits.

Well, perhaps we can draw our conclusions on what we might need to do.

I think that is the only legal way to do it, I am afraid.

Thank you, Mr Hughes.

Just to explain the £12 billion number in the report that you referred to, in our forecast for incapacity and disability benefits, we have seen a huge increase in caseload since covid. There are about 1 million extra people on the caseload, and about £20 billion of extra spending. In the baseline forecast in March, even including the policy changes that the Government had announced at that time, spending was still forecast to increase by a further £15 billion—so still a big pressure from those benefits on the public finances. That forecast was based on an assumption—a judgment—we had made that some of the growth in the caseload we have seen since covid has essentially been due to cost of living pressures pushing people into the more generous benefits. We made a judgment that once, in our forecast, the cost of living pressures ease off, some of that caseload growth will also ease off. The £12 billion in the report was the estimate of the risk if that judgment turns out to be wrong, and caseload actually increases at the same rate we have since covid—an extra £12 billion of spending. It is definitely worth saying that one of the reasons that we highlighted that risk is that the drivers of the big increase in cases that we have seen are quite uncertain. It is probably a combination of changes in the underlying health of the population, the working-age population deteriorating, and changes in the labour market and cost of living pressures, but also structural factors in the benefits system itself.

On that specific point, do you identify challenges derived from the different pathway for mental health as a specific particular challenge in terms of the volume and cost of welfare?

It is difficult to isolate the specific costs of that driver, but we have certainly highlighted it in some of the analysis we have done as one of the most important drivers of the additional caseload from increasing mental health issues, especially among younger people.

When I read your fiscal risks and sustainability report this time, I immediately thought of that Hemingway quote about how you go bankrupt: in two ways—gradually, then suddenly. I thought that the report looked like we had entered the “suddenly” phase. I wanted to ask you about one of the things that you scored in your 2024 report that appears to be suddenly changing. That is the very high risk you allocated to scoring the change to the taxation of non-domiciled residents in the UK. You put it in the category of there being really high uncertainty around it, regarding behavioural changes. You scored specifically the change to charging inheritance tax on worldwide assets as being about 0.1 in terms of revenue over the next few years. It seems to be one of those things, according to media reports, that is suddenly changing behaviour. Does the Office for Budget Responsibility have any way of pulling a ripcord and saying, “Help. We think we got this one wrong. Our central assumption is clearly not right. This is going to cost the Treasury a lot more than we thought. Please do something”?

You are absolutely right: we have highlighted that this is a very uncertain costing. It affects a relatively small group of taxpayers and ones who are likely to have a large behavioural response to tax changes. We will not update the costing of that until the next fiscal event, and at the moment, there is no hard evidence on the impact that it is having. It is quite likely that there will not be any hard evidence for quite some time until we actually get firm out-turn data from—

From HMRC?

Yes, from HMRC. In the reforms that the previous Government made in their final fiscal events, we assumed that there would be a migration response to the non-doms regime changes of around 10% to 20%—so 10% to 20% of non-doms would leave the country as a result of those reforms. We then changed that assumption in October, on the basis of the changes that the current Government made, particularly in respect of inheritance tax. We increased it to about 12% to 25% of non-doms leaving as a result of the reforms. So quite a substantial migration assumption is already assumed in the costing. The 25% essentially refers to the very highest wealth individuals—ones who have trusts. The 12% refers to those without trusts. So a pretty big migration assumption is already baked into that costing. As I say, I do not think we will really have any hard evidence on the actual outcome for quite some time.

What I am hearing you say is that they could all have left by the time you get the actual numbers.

Well, no—what I am saying is that we will not have any hard data on the actual numbers who have left until we get the tax returns. Even at that point, it is probably going to be quite difficult to isolate the impacts of this specific policy change against lots of other factors that affect location decisions from this group, who are obviously a highly mobile group of individuals.

What comparators would you use, though, for a normal year with non-doms, because they are highly mobile people, potentially with assets all over the world? Can you isolate the impacts of some of these changes?

We will try very hard and work with HMRC to try to isolate the impacts of these policies. However, as I say, there is a lot of turnover in this population anyway, so it will likely be quite difficult. We based those migration assumptions on the behaviour we saw after the previous set of reforms in 2017, when, actually, the response was less large than quite a lot of people were expecting.

What you scored for—anecdotally—the inheritance tax changes, which seem to be having the biggest impact in terms of people’s location decisions, was 0.1 every year, in terms of the net figure. So it is not a huge revenue inbound, but you do not have any way that you can pull the emergency cord and warn the Treasury about these assumptions potentially being tested.

As I say, we do our costings updates at fiscal events. At that point, we will assess whatever evidence there is.

I have a very quick question, Mr Hughes, about the climate risks that you point to. May I also just commend the OBR on its work on this issue, because it has been incredibly useful in the debate, as well as in setting out the investments and the costs? From where I am sitting, it looks like we are having a lot of damage going forward accounted for in the costs of transition. But all the productivity benefits of the transition—investment in skills, clean energy, reducing the marginal price and more capital per worker—do not seem to be included in the productivity estimates of future growth. Is that a fair reading of where this is at the moment?

Tom might want to say more, but I can have a go first. When we look at the cost of net zero transition and climate damage, generally speaking, the cost of climate damage seems to be going up while net zero seems to be getting cheaper, relative to our previous estimates and for a whole host of reasons. In terms of whether you could make the net zero transition even cheaper by getting some sort of economic productivity dividend, we do not assume that in how we do our projections. That is for a number of reasons, including partly because no serious climate economist can give us a number, and we are definitely not climate economists in the OBR. We rely an awful lot on empirical estimates that come from the literature and, in particular, people who can do it for the UK economy—not for some other economy with a different configuration of energy supply and other things. Secondly, if you think back, we are more or less leading the charge in carbon reduction in the UK—and have been for decades—and our productivity has been going down. There is not an obvious correlation in recent history between making the transition to net zero and boosting productivity. I am not saying that one is the cause of the other, but what we have not seen is our being ahead of other countries in the net zero transition putting us ahead of them in the great race to make our economies productive again. That relationship does not emerge obviously from the data. We have asked others who are much more expert in climate economics than us, “If there were a fair number to put into our forecasts or even into some scenarios, what would that number be?” There are people on either side of that debate. There are people who say that the net zero transition is going to be good for productivity, because it is going to make the unit cost of energy lower. There are other people on the other side of the argument who say it is going to make the economy less productive, because for a period, you are raising the unit cost of energy that you currently use because it still mostly comes from fossil fuels, and that will hurt energy-intensive businesses and squeeze margins in any business that has to keep the lights on, and that is going to have an adverse effect on their investment in other things. In the end, we have ended up in an unsatisfactory neutral position, where we do not know which side of that debate is going to turn out to be right. For the moment, nobody has made a persuasive case one way or the other, and there is not much in history that can give you a positive correlation between reducing your emissions and boosting productivity.

Is that not part of the worry? This gets to the wider point about non-quantifiable benefits. In the post-war period, there was huge investment that led to huge productivity gains at that moment. If at that point we had the OBR, it would be very difficult to see how we could have invested in anything because people would have said, “Well, there is not going to be a benefit.” To the point about energy costs, I would like to ask for a slight clarification. I am a little confused when you say that this will not improve productivity. You are right that at the moment we have built out on clean energy on one side, but on the other, we still have a lot of natural gas, which is therefore always setting the marginal price. The reason why we have such high industrial energy prices is precisely that natural gas is setting the marginal price. If we accept that if you get off natural gas for some of your marginal pricing, you are going to have lower energy bills, would that not lead to a productivity improvement?

It depends. Part of it depends on how much you need to pay for storage, because the problem with wind and solar is that it is intermittent. So how much are you paying for storage and what does that do to the unit cost? If you are getting your baseload from nuclear, we know nuclear is very expensive and can drive up the unit cost of energy system-wide. And then, in other areas, like heating, you can switch your gas boiler to a heat pump that heats your house in a less dirty way but does not do it any better—it just keeps your house warm. It is not obvious that you are getting a disposable income benefit from that transition either, and people are spending a lot of money to switch from one to the other. On this particular question, that is why we have ended up with a neutral position. It is not correct to characterise us as saying that we do not believe in the growth effects of public investment. We do, and exactly as we discussed earlier, that is reflected in our forecasts.

Another place where we will suddenly start losing revenue is fuel duty. In your current numbers and projection, when does the temporary 5p reduction in fuel duty end?

I believe that it ends next spring.

Yes, and when will the inflection point be where it has such a major impact on the number of petrol cars on the road that Chancellors are going to have to step in and do something?

As you mentioned, in our analysis of the long-term costs for Government of the transition to net zero, the loss of fuel duty is essentially the biggest. It amounts to around two thirds of the total cost.

Clearly, a Chancellor will have to do something about that.

We have a scenario in the report that looks at how you could reduce the costs of that transition, depending on the policy path you choose—the decision on policy path is obviously up to Government. To illustrate the risk, if the Government spend less on the transition and choose a policy path in which they use other tools, such as regulation or taxation, and they replace fuel duty with potentially another motoring tax, it almost halves the cost from the impact on debt of the transition to net zero in our scenarios. It is also the case, even over the medium-term forecast, that you can start to see the impacts of the loss of revenue from electric vehicles. Even with the assumption that we have to include in our forecast that fuel duty rises in line with RPI, by the end of the forecast, fuel duty starts to fall as a share of GDP, because of increased electrification.

What year is that, just for my understanding?

From memory, it is right at the end of the current forecast.

What year is that?

It is 2029-30.

But all other things being equal, if the Chancellor does not make a decision in this year’s fiscal event, fuel duty will go up 5p next spring.

That is the assumption in the forecast.

There is a very small unit in the Treasury that is looking at some of these green taxes. Do you have any engagement with them? It is quite a small number of people.

Do we have any engagement with the unit in the Treasury? Sorry, I missed what you said.

You obviously only discuss actual Government policy, but there is a team in the Treasury looking at the issue of what is a green tax and how you deal with the fuel duty issue. Do you engage with them?

I see. On principle, we do not discuss our engagement with Treasury on policy issues ahead of fiscal events.

Fine, okay—it was a good try.

Mr Josephs, let us imagine that you have come with me to meet a bunch of people in their late 20s or early 30s in my constituency to talk about pensions. They are all telling you that you have the state pension, and they are putting a bit into their pension, but the self-employed couple in the corner are not. Could you talk us through what you see as the main problems with our current pension system, from the point of view of both the sustainability of public finances and securing a comfortable retirement for that group of 20 and 30-year-olds?

The risks we highlight for the long-term sustainability of the pension system are driven by three things. We look at state pensions, where the demographic trends and the ageing population mean that, on the basis of current policy, spending will increase over time by about 1.5% of GDP. That is the second biggest source of upward pressure on spending over the long term, after healthcare. We also look at the risks from the triple lock, which we show is not only a source of pressure but a source of risk. If we have a scenario similar to the one we have seen since 2010—where inflation has been much more volatile than previously, and earnings have been lower—and project that out, the triple lock adds, on top of the baseline projection, around another 1.5% of GDP of pressure on the public finances. That is a substantial pressure on the public finances over that timeframe. The second risk that we look at goes to your point on self-employed people. The transition from DB to DC pensions and auto-enrolment has meant that a lot more people are now saving for a private pension than did previously. Sorry, I have now moved from the state pension to private pensions. A lot more people are saving in private pensions, but there is quite a lot of evidence—here we drew on studies by the IFS, the DWP and others—that a number of groups look as if they might not be saving enough to give them an adequate pension in retirement. Particular groups, and the self-employed are one, are not taking up private pensions to the same degree as others. That is not necessarily a direct risk to the public finances, but it creates some risk that the Government, in the future, might have to provide more support to those groups when they reach retirement. The final risk we highlight is the decline of the DB sector, which we talked about earlier.

With the decline in DB and people moving to DC, they bear the investment risk. In your report, you identified housing benefit as one factor that is driving up the non-pension benefits bill for pensioners, because home ownership has dropped. Presumably another risk is that, with the switch from DB to DC, people are not saving enough, so there is an additional pressure on the Government budget to ensure that people have a reasonable minimum living standard.

You are right. That is another group we highlight in the report, again drawing on the IFS analysis showing that people in private rented accommodation are projected to have less adequate retirement incomes than other groups. That could create pressure on the housing benefit bill for future Governments, or just pressure to provide wider support in the future.

In 2070, you show 7.7% or so going on pensions and 1.3% going on pensioner benefits, including housing benefits. That gets us to 9% or so of GDP being spent on pensions and pensioner benefits. Is that sustainable?

The overall picture we present with our long-term projections is one of debt on an unsustainable path. The main contributor is that, on the basis of current policy settings and demographic assumptions, a number of elements put upward pressure on spending, while tax as a share of GDP is projected to be broadly stable, or slightly falling because of the loss of fuel duty. A combination of spending pressures from a number of sources, of which pensions are a big one, creates that gap.

In that combination, how material is the pensions element?

After healthcare, it is the second biggest contributor to the increase in spending over our long-term projections.

I suspect you will politely decline the offer to set out some options that the Government could look at to make this more sustainable.

Again, we are not here to comment on policy, but we provide different scenarios in the report for the impact of the different factors that contribute to this increase. Demographics is one, and the triple lock is obviously one. We also look at the impact of increases in the state pension age, and over the long term, that actually makes quite a considerable difference in reducing the pressure on state pension spending over time. It lowers it by about 1% of GDP in total. We assume three increases in the state pension age over that 50-year projection.

Thank you very much. That is certainly an important long-term issue. That concludes this part of our hearing.