Foreword

This Economic and fiscal outlook (EFO) sets out our central forecast and the uncertainties that surround it for the five years to 2028-29, taking account of recent data and government policies announced up to and including the Spring Budget 2024. The forecasts presented in this document represent our collective view as the three independent members of the OBR’s Budget Responsibility Committee (BRC). We take full responsibility for the judgements that underpin them and for the conclusions we have reached.

As always, we have been greatly supported in our work by the staff of the OBR. We are very grateful for their commitment, professionalism, and expertise. We have also drawn on the analysis of numerous officials across government in preparing these forecasts, including in HM Treasury, HM Revenue and Customs, the Department for Work and Pensions, the Department for Levelling Up, Housing and Communities, the Department for Education, the Department for Energy Security and Net Zero, the Home Office, the Department for Transport, the Department of Health and Social Care, Department for Digital, Culture, Media and Sport, North Sea Transition Authority, the Office for National Statistics, the UK Debt Management Office, National Savings and Investments, the British Business Bank, the BBC, Homes England, UK Government Investments, the Government Actuary’s Department, the Insolvency Service, the Scottish Government, the Scottish Fiscal Commission, the Welsh Government, the Department for Communities and the Department of Finance in Northern Ireland, Transport for London, and various public service pension schemes. We also held helpful discussions with various departmental finance directors. We are grateful to all of them for their engagement and insights.

Outside government we have held useful discussions with the Bank of England, National Institute of Economic and Social Research, the Institute for Fiscal Studies, Resolution Foundation and Institute for Government. We have also benefitted from discussions with: Hamish Low, Madeleine Sumption, Alan Manning and Tessa Hall. We are also grateful to our labour market and migration, and macroeconomics and econometrics, advisory panel subgroups for sharing their insights.

The date for the Budget and this forecast was announced on 27 December, giving exactly the required ten weeks’ notice for fiscal events specified in the Memorandum of understanding between the Office for Budget Responsibility, HM Treasury, the Department for Work and Pensions and HM Revenue and Customs (MoU). But the earlier-than-usual March Budget date, and the publication of key ONS population and labour market releases in early February, effectively constrained the window for the pre-measures forecast process to five weeks. As a result, we only completed two, rather than the usual three, rounds of pre-measures forecasts. This made the forecast process more challenging than usual, but it has not impaired the final forecast.

We published the timetable of the key stages of the forecast on 25 January, once it had been agreed by signatories of the MoU. On this occasion, the timetable was adjusted to permit 23 working days instead of 21 between our final pre-measures economy forecast and the Budget date. This was to accommodate a Treasury request to extend the initial window for major Budget policy decisions from five to six days following the delivery of our final pre-measures fiscal forecast. The window for the preparation of our pre-measures economy forecast was also extended by a day, to give sufficient time for us to consider fully the implications of the latest ONS data. We are grateful to ONS for providing us with pre-release access to these datasets, which made this forecast timetable feasible. Within the constraints of this compressed timetable, the overall process operated effectively and deadlines were adhered to at each stage. The forecast timetable for this EFO has been:

• OBR staff prepared an initial economy forecast, incorporating latest interest rates and other market determinants and data released since our previous forecast in November 2023. This first economy forecast was sent to the Chancellor on 17 January.
• Using the economic determinants from this forecast (such as the components of nominal income and spending, unemployment, inflation, and interest rates), we commissioned forecasts from the relevant government departments for the various tax and spending items that in aggregate determine the position of the public finances. We discussed these in detail with the officials producing them, which allowed us to investigate proposed changes in forecasting methodology and to assess the significance of recent tax and spending outturn data. In many cases the BRC requested changes to methodology and/or the interpretation of recent data. This first fiscal forecast was sent to the Chancellor on 30 January.
• We then produced a second and final pre-measures economy forecast, in which we took on the latest ONS population projections and reweighted LFS data, and incorporated judgements embodied in our first fiscal forecast. This economy forecast included energy and financial market data based on the average over the 10 working days to 23 January and was sent to the Treasury on 5 February. Where we thought it would be helpful, we commissioned analysis from the relevant teams in the Treasury to inform our views and discussed general forecast issues, though not specific judgements, with experts from external organisations.
• This second economy forecast provided the basis for the next round of fiscal forecasts. Discussions with HMRC, DWP and other departments gave us the opportunity to consider further analysis, methodological changes, and alternative forecast judgements. We sent our second and final pre-measures fiscal forecast to the Chancellor on 14 February.
• In parallel, we undertook a process of engagement and analysis to assess the set of policy measures to be announced in the Budget that we deemed could have specific effects on our economy forecast, to inform our forecast judgements. This involved several rounds of engagement with the Treasury and other departments as both the specification of policy packages and our assessment of their impact were refined.
• We also scrutinised the costing of individual tax and spending measures announced since the November 2023 forecast. As usual, OBR staff and the BRC requested further information and/or changes to almost all the draft costings prepared by departments. We have certified all policy measures in the forecast as reasonable and central.
• Alongside the development of the final economy forecast we made an initial assessment of the economic and fiscal effects of the emerging policy package. This built on earlier analysis that allowed us to factor in an initial package of measures that was provided by the Treasury on 16 February. We incorporated this package into a preliminary post-measures forecast, in order to provide an early view on the effect of Budget measures on the economy and public finances, which we sent to the Chancellor on 20 February. This forecast round was produced using our internal ready-reckoner models (rather than being sent to departmental forecasters).
• In line with the agreed timetable, on 22 February the Treasury provided the final package of measures that we judged would cause movements in our economy forecast. We sent the resulting final economy forecast to the Treasury on 27 February and a near-final fiscal forecast on 28 February. Final policy decisions, that would not be material for the economy forecast, were provided by the Treasury on 29 February and our fiscal forecast was then finalised on 1 March and sent to the Treasury on the same day.
• The Treasury made a written request, as provided for in the MoU between us, that we provide the Chancellor and an agreed list of his special advisers and officials with a near-final draft of the EFO on 1 March. This allowed the Treasury to prepare the Chancellor’s statement. We also provided pre-release access to the full and final EFO on 4 March.

During the forecasting period, the BRC held dozens of scrutiny and challenge meetings with officials from other departments, in addition to numerous further meetings at staff level and those with external stakeholders. We have been provided with all the information and analysis that we requested and have come under no pressure from Ministers, advisers, or officials to change any of our conclusions as the forecast has progressed. The BRC met with the Chancellor on three occasions to discuss the forecast over the course of its production (on 1 February, 13 February and 27 February). A full log of our substantive contact with Ministers, their offices and special advisers can be found on our website. This includes the list of special advisers and officials who received the near-final draft of the EFO on 1 March.

Our non-executive members, Bronwyn Curtis OBE, Baroness Hogg, and Dame Susan Rice provide additional assurance over how we engage with the Treasury and other departments. This includes reviewing any correspondence that OBR staff feel either breaches the MoU requirement that it be confined to factual comments only or could be construed as doing so. That review takes places as soon as practicable after each EFO has been published. Any concerns our non-executive members have will be raised with the Treasury’s Permanent Secretary or the Treasury Select Committee if they deem that appropriate.

We would be pleased to receive feedback on any aspect of the content or presentation of our analysis. This can be sent to [email protected].

The Budget Responsibility Committee

Richard Hughes, Professor David Miles CBE and Tom Josephs

Chapter 1: Executive summary

Overview

1.1 The UK economy has emerged from the twin global shocks of the pandemic and Russian invasion of Ukraine into a period of declining inflation but stagnating output. Inflation has receded more quickly than we expected in November and markets now expect a sharper decline in interest rates. This strengthens near-term growth prospects and should enable a faster recovery in living standards from last financial year’s record decline. But the medium-term economic outlook remains challenging. One of the biggest changes to our economy forecast is an increase in the size and growth of the UK population. But higher and rising levels of inactivity offset its impact on the overall size of the workforce, leaving our forecast for the level of GDP in five years virtually unchanged from the autumn, and the level of GDP per person slightly lower.

1.2 The overall outlook for the public finances is also similar to November. Lower inflation and interest rates reduce the Government’s projected debt servicing and welfare costs, but they also reduce revenues. These pre-measures forecast changes result in a £20 billion fiscal improvement over the next two years but leave borrowing largely unchanged in five years’ time. The Budget announces a package of net tax cuts, including a further 2p cut to the main rates of employee and self-employed national insurance contributions, the cost of which is partially recouped by tax rises in later years. Borrowing is still projected to fall in each of the next five years thanks to tax as a share of GDP rising to near to a post-war high, debt interest costs falling, and per person spending on public services being held flat in real terms. This is just enough to meet the Government’s fiscal rules on our central forecast with underlying debt falling as a share of GDP in 2028-29 by a historically modest margin of £8.9 billion.

1.3 This margin is a small fraction of the risks around that central forecast. Inflation could rebound and remain higher for longer if the conflict in the Middle East were to widen or if domestic wage pressures do not subside as quickly as we assume. There is also uncertainty around several key drivers of medium-term economic growth including net migration, labour market participation, and productivity growth. The fiscal forecast is highly sensitive to movements in interest rates which have been unusually volatile recently. Since November, market expectations for medium-term Bank Rate have been both 1 percentage point higher but also ½ a percentage point lower than assumed in this forecast. The fiscal forecast is also conditioned on the tax take rising to near record highs, including through planned rises in fuel duty that have not, in practice, been implemented since 2011. It also assumes the Government will stick to assumptions which imply no real growth in public spending per person over the next five years, despite committing to increase spending on some major public services in line with or faster than GDP.

Economic outlook

1.4 CPI inflation was 4.2 per cent in the final quarter of last year, 0.6 percentage points lower than we forecast in November. We now expect it to fall further to an average of 2.2 per cent this year and 1.5 per cent in 2025 before gradually returning to target at the end of the forecast period. Our lower central forecast for inflation is partly driven by larger anticipated falls in global energy prices. We also expect domestically generated inflation to be weaker, as falling energy prices pass through into lower economy-wide costs and the labour market continues to loosen. Our central forecast assumes current disruptions in the Red Sea make only a small (0.2 percentage points) upward contribution to inflation. But we also consider the risks of a widening conflict in the Middle East, through a scenario in which a sharp rise in energy prices causes inflation to spike back up to an annual peak of almost 6 per cent.

Chart 1.1: CPI inflation

1.5 Alongside easing inflationary pressures, market participants now expect a sharper fall in interest rates than in the autumn. Bank Rate is expected to fall more steeply this year from its current peak of 5.25 per cent to 4.2 per cent in the final quarter of 2024. In the medium term, Bank Rate falls further to 3.3 per cent, almost ¾ of a percentage point lower than in our November forecast. With gilt yields also around ½ a percentage point lower across maturities, the cost of debt interest for the Government is significantly lower than expected in November. But expectations remain volatile, as shown by expectations for Bank Rate in 2028 oscillating between 2.7 and 4.2 per cent since our November forecast.

1.6 There has been important news since November about the size and projected growth of the UK population. Based on updated outturn data and January ONS population projections, we now expect the total UK adult population to rise from 55 million in 2023 to 57 million by the end of the forecast, 1¾ per cent larger (or 1 million more people) than in November. About two thirds of this increase is due to a higher estimate of the current UK population, which takes account of the 2021 Census and net migration since then, with the remainder largely from higher net migration over the forecast. Having reached a record high of 745,000 in 2022, net migration stood at 670,000 in the year to mid-2023, around 70,000 higher than assumed in our November forecast. Supported by the policy measures announced since our last forecast, our central forecast assumes annual net migration falls back to 315,000 in the medium term, up from 245,000 in our November forecast. We also explore scenarios in which annual net migration is around 200,000 higher or lower, which could raise or lower the level of GDP in 2028-29 by around 1½ per cent but would have a small impact of uncertain direction on GDP per person.

1.7 The latest data also suggest that the post-pandemic rise in economic inactivity is likely to prove more persistent than we previously thought. The number of inactive working-age adults is no longer declining from its post-pandemic peak, as previous data suggested, and has instead rebounded to 9.3 million. This keeps it around its highest level in over a decade and 700,000 more than before the pandemic. Around one third of the working-age inactive population cite long-term illness as their principal reason for not being in the labour force. We estimate that the policies on childcare expansion, welfare reform, and personal tax cuts announced over the past three fiscal events will increase total labour supply by over 300,000 people in full-time equivalent terms. But the ongoing ‘fiscal drag’ from frozen personal tax thresholds will also weigh on work incentives, offsetting over a third of this rise for an overall change of close to 200,000. But after taking account of all of the factors discussed above, we now expect the labour participation rate to continue falling from its pre-pandemic quarterly peak of 64.3 per cent to 62.8 per cent by 2028, 0.5 percentage points below our November forecast.

Chart 1.2: Net migration and participation rate

1.8 With a larger population but lower labour participation, our forecast for potential output growth over the next five years is largely unchanged from November at around 1⅔ per cent a year. Higher net migration, lower interest rates, and lower energy prices boost population growth, business investment, and productivity respectively. But the latest data on labour participation, demographic and other factors have led us to revise down the overall trend participation rate and average hours worked. The net effect of these changes leaves the level of output 0.1 per cent lower in 2028 than forecast in November, but 0.1 per cent higher after accounting for the policies in this Budget that boost labour supply.

1.9 The profile of output is slightly weaker in the near term but slightly stronger in the latter part of the decade. GDP grew by only 0.1 per cent in 2023, undershooting our November forecast by 0.4 percentage points. But we expect output growth to pick up to 0.8 per cent in 2024 as interest rates fall and real household incomes recover. GDP growth picks up to around 2 per cent in the middle of the decade as slack in the economy is taken up, before falling back towards its assumed trend rate of around 1⅔ per cent by 2028. Policies announced in this Spring Budget provide a small temporary boost to demand in the near term and to supply in the medium term which raises the level of real GDP by 0.2 per cent in 2028-29. Risks to our medium-term real GDP forecast remain elevated. The outlook for productivity growth is our most important and uncertain forecast judgement, and there is significant uncertainty over both our migration and participation forecasts.

Chart 1.3: Real GDP and real GDP per person

1.10 Having steadily declined since early 2022, real GDP per person is forecast to trough at 1¼ per cent below its pre-pandemic peak in the first half of 2024. Persistent weakness in per person output has been driven by rises in inactivity and subdued productivity growth, which has remained well below its pre-financial crisis average in recent years, even after accounting for the rebound from the pandemic. We expect real GDP per person to begin to recover later this year and regain its pre-pandemic level in 2025. Real GDP per person increases to around 4½ per cent above its pre-pandemic peak by the forecast horizon, but remains around ¾ of a per cent lower than we forecast in November.

1.11 The recovery in output is largely driven by a pick-up in household consumption growth to around 2 per cent from 2025 to 2028. That growth is stronger than in November due to higher household disposable incomes, a sharper slowdown in inflation, and lower interest rates. In contrast, business investment is expected to contract by around 5 per cent this year as past increases in interest rates raise the cost of capital and weigh on capital spending. Investment then strengthens as interest rates fall and demand picks up. Recent trade data have been volatile and subject to large revisions. Nevertheless, we forecast that trade volumes will continue to be subdued in the next few years due to sluggish growth in the UK and global economies, and the evolving impact of Brexit. From 2024 to 2028, we expect export and import volumes to average growth of 0.3 per cent and 0.1 per cent a year, respectively. As a result, net trade makes a negligible contribution to growth over the forecast period.

1.12 Weak near-term GDP growth drives a modest rise in the unemployment rate in 2024, which then falls back as the recovery gathers pace. The latest ONS data suggest that unemployment fell to 3.8 per cent in the fourth quarter of 2023. In contrast, claimant count data measuring the number of people on unemployment benefits has remained flat in recent months, while the redundancy rate has been on a slow upwards trend since the middle of 2022. We therefore judge that this and wider evidence is consistent with a moderate rise in the unemployment rate, peaking at 4.5 per cent in the last quarter of 2024, in line with our forecast for subdued economic growth and increasing spare capacity in the economy. The peak in unemployment of around 1.6 million people is marginally lower (by about 40,000 people) than in our November forecast, though it comes half a year sooner. The unemployment rate is then forecast to decline to its estimated structural level of 4.1 per cent by 2028.

1.13 From a thirty-year high of close to 7 per cent, we expect nominal average pay growth to halve in 2024 as inflation falls and the labour market loosens. Nominal earnings growth is slightly weaker than our November forecast as inflation is less persistent. In terms of total labour income, this weaker nominal earnings growth is partly offset by faster growth in the number of employees. That leaves the level of nominal wages and salaries, a key determinant of our fiscal forecast, 0.3 per cent lower at the forecast horizon than in November. This is broadly in line with our revision to nominal GDP.

1.14 Living standards are expected to recover more quickly than we forecast in November and grow by around 1 per cent a year on average over the forecast. 2022-23 remains the fiscal year with the largest year-on-year drop in living standards since ONS records began in the 1950s. But we now forecast real household disposable income per person to recover its pre-pandemic peak by 2025-26, two years earlier than in our November forecast. The faster recovery in living standards arises because the negative terms of trade shock brought about by the rise in the price of imported energy has unwound more quickly and fully than expected. Policies in this Budget provide an additional boost to household incomes, with the further reduction in the main rates of national insurance contributions alone providing a direct boost of ½ a per cent.

Chart 1.4: Real household disposable income per person

1.15 Nominal GDP is forecast to grow by an average of 3.3 per cent a year over the forecast period, slightly less than expected in November due to the lower inflation outlook. The slower growth in the GDP deflator, combined with a little-changed path for real GDP, leaves the level of nominal GDP 0.3 per cent lower in the medium term than in our last forecast. In terms of the other key nominal tax bases, both profit and nominal consumption growth are weaker in the near term but rise faster in the medium term. By the forecast horizon, the change in both tax bases is broadly in line with that of nominal GDP.

Fiscal outlook

1.16 The fiscal position remains very challenging due to high debt, subdued economic growth, and the highest interest rates for over a decade. Economic and market conditions mean the Government needs to run a primary surplus (revenues minus spending, excluding interest) of around 1.3 per cent of GDP just to stabilise debt in the medium term. By comparison across the 2010s, on average, debt could be stabilised while running a primary deficit of 2.1 per cent. Based on its current policies, we forecast that the Government achieves a primary surplus and falling debt by 2028-29, through a 1.1 per cent of GDP increase in primary receipts, and a 1.7 per cent fall in primary spending as a share of GDP. However, the fiscal deficit is the difference between two very large numbers (total revenues and total spending both now exceed £1 trillion), and the uncertainty around our central forecast for the trajectory of both borrowing and debt over the next five years is substantial.

1.17 Relative to our November forecast, the underlying pre-measures fiscal outlook is slightly better in the near term but largely unchanged by the end of the forecast period. Borrowing is expected to be £10.1 billion lower this year and around £10 billion lower over each of the next two years. This is largely due to lower interest rates and inflation which reduce debt interest and welfare costs. But lower inflation also lowers nominal tax receipts by increasing amounts over the forecast. Overall, before taking account of Budget policy measures, this leaves borrowing in 2028-29 only £0.8 billion lower than we forecast in November.

1.18 Against this fiscal backdrop, the Government has announced a frontloaded package of measures, including significant net tax cuts, estimated to increase borrowing on average by £8.0 billion a year. Overall, the direct and indirect effects of the measures raise borrowing by £12.7 billion in 2024-25, tapering down to £5.2 billion in 2028-29. This results in slightly higher post-measures borrowing in every year of the five-year forecast period than in November, peaking at £4.4 billion more in 2028-29.

Chart 1.5: Public sector net borrowing: change since November

1.19 The Budget measures with the largest estimated direct fiscal impact are:

• a further reduction in rates of national insurance contributions (NICs), including a 2p cut in the main rate of employee and self-employed NICs from April 2024, which costs £10.7 billion by 2028-29;
• reform of the current non-domicile regime from April 2025, which raises £3.1 billion on average from 2026-27 to 2028-29;
• a number of new taxes and other revenue raising measures, including the introduction of vaping duty and the carbon border adjustment mechanism, HMRC anti-avoidance and compliance measures, and a one-year extension to the energy profits levy, which collectively raise £3.9 billion by 2028-29; and
• £0.9 billion more departmental capital spending per year on average between 2025-26 and 2027-28 on a public sector productivity programme focused on the NHS, and £0.8 billion per year less departmental resource spending from 2025-26 onwards.

1.20 The indirect effects of these policies on the economy deliver a modest boost to demand in the near term and to labour supply in the medium term.
The demand boost is primarily driven by the increase to household incomes provided by the near-term tax-cuts. The cut in NICs and reform to the high-income child benefit taper also increase labour supply by improving work incentives. By the end of the forecast around half of the fiscal benefit from these effects is offset by the additional debt interest spending on the borrowing required to fund the policy package.

1.21 Tax as a share of GDP is forecast to rise to 37.1 per cent of GDP in 2028-29, 4.0 per cent of GDP higher than the pre-pandemic level. Of this rise, we estimate 2.9 per cent will have taken place by the end of 2023-24, with the remaining 1.1 per cent forecast to take place thereafter to 2028-29. Over the whole period, major tax-raising policy decisions since March 2020 raise the tax take by 2.0 per cent of GDP, of which the freezes to income tax thresholds account for 1.3 per cent of GDP. These are partially offset by major tax cuts amounting to 0.9 per cent of GDP, of which the cuts to NICs announced in this Budget and in November 2023 account for 0.7 per cent of GDP.

1.22 Spending is forecast to fall steadily as a share of GDP from 44.5 per cent this year to 42.5 per cent in 2028-29, but remains around 3 per cent of GDP above its pre-pandemic level. Debt interest costs account for around 0.4 per cent of GDP of the reduction from this year to 2028-29. The single largest contribution to the decline comes from the assumed 1.0 per cent of GDP reduction in departmental spending on public services over the next five years. Partially offsetting this, welfare spending is forecast to increase by 0.4 per cent of GDP as rising caseloads drive higher spending on health and disability-related benefits, and a combination of an ageing population and the triple lock drive higher pension spending.

1.23 In our central forecast, borrowing as a share of GDP is expected to fall steadily from 4.2 per cent this year to 1.2 per cent of GDP in 2028-29. Of this 2.9 per cent of GDP reduction, the forecast rise in receipts contributes 0.9 per cent of GDP, while the forecast decline in spending contributes 2.1 per cent of GDP. In cash terms, borrowing is forecast to fall from £114.1 billion in 2023-24 to £39.4 billion in 2028-29. Compared to our November forecast, borrowing is slightly higher by 0.1 per cent of GDP on average over the five-year forecast period. As illustrated by Chart 1.6, there are a wide range of risks around our central borrowing forecast, which are discussed further below.

Chart 1.6: Public sector net borrowing

1.24 Public sector net debt (excluding the Bank of England) in the central forecast rises from 88.8 per cent of GDP this year to a peak of 93.2 per cent of GDP in 2027-28, before falling slightly to 92.9 per cent of GDP in 2028-29. This is a very similar path to that in our previous forecast. The wider measure of headline debt rises from 97.6 per cent of GDP this year to 98.8 per cent in 2024-25, then falls to 94.3 per cent of GDP in 2028-29.

Chart 1.7: Public sector net debt excluding the Bank of England

Performance against the Government’s fiscal targets

1.25 The Government’s primary fiscal target is for public sector net debt excluding the Bank of England to fall in the fifth, and final, year of the forecast. In our central forecast, this rule is met by a margin of £8.9 billion (0.3 per cent of GDP), down from the £13.0 billion (0.4 per cent of GDP) margin in our November forecast. As the pre-measures forecast is broadly unchanged by 2028-29, the reduction in headroom is mostly due to the impact of policy measures.

1.26 Headroom of £8.9 billion is lower than the average of £26.1 billion that Chancellors have set aside against their fiscal rules since 2010.
And it is very low compared to the risks around our forecast, which are summarised below. Based on historical forecast errors at the five-year horizon, we estimate there is a 54 per cent probability that the fiscal target would be met on current policy. This compares to a 56 per cent probability in our November forecast. Our forecast incorporates £4.8 billion of extra revenue in 2028-29 from the Government’s stated policy of reversing the 5p cut and increasing fuel duty rates in line with RPI inflation. If, as at this Budget and at all fiscal events since 2011, duty rates were instead held at the current rate, then almost half of the headroom in 2028-29 would be removed.

Chart 1.8: Underlying debt falling headroom: changes since November

1.27 The supplementary target of public sector net borrowing being below 3 per cent of GDP in 2028-29 is also met in our central forecast. It is met by a larger margin than the fiscal mandate: £56.8 billion, or 1.8 per cent of GDP (down by £4.8 billion since November). History suggests that this margin would be consistent with a 72 per cent chance of meeting the supplementary target.

Risks and uncertainties

1.28 Historically large changes in energy prices, interest rates, wage growth, and population growth have driven significant revisions to our recent economic and fiscal forecasts. We continue to emphasise the uncertainties around our forecasts and the possibility that any of our key judgements could prove too optimistic or pessimistic. Key risks for this forecast include:

• While our central forecast is for inflation to return to target this year, the outlook remains highly uncertain. Externally, conflict in the Middle East poses risks to global goods and energy markets. The scenario in Box 2.2 illustrates how a widening of the current conflict in the Middle East could significantly reduce energy supply from the region and further disrupt global supply chains for goods. In this scenario, quarterly inflation spikes back up to over 7 per cent. We estimate that this could raise borrowing by £23.1 billion on average over the five-year forecast and leave underlying debt 0.8 per cent higher as a share of GDP by 2028-29. Domestically, our central projection is for wage growth to slow over the coming year, but wages may prove stickier and keep inflation higher for longer.
• Our forecast is based on market expectations for Bank Rate and gilt yields, and, while these have fallen significantly as inflation has dropped, they remain unusually volatile (see Box 4.3). Since we closed our November forecast, expectations for medium-term Bank Rate have moved between 2.7 and 4.2 per cent and 10-year gilt spot yields have oscillated between 3.5 and 4.7 per cent. If effective interest rates on all central government debt were just 0.3 percentage points higher or lower, this would eliminate the headroom to debt falling in 2028-29, or increase it by around £9 billion, respectively.
• A major change in our forecast since November has been incorporating updated net migration projections and Labour Force Survey data from the ONS. Future migration levels are highly uncertain and difficult to forecast, particularly given policy changes announced since November. Our scenarios show that if annual net migration was around 200,000 higher or lower than the ONS projection of 315,000 in the medium term, it might raise or lower GDP by around 1½ per cent in 2028-29, respectively. But the impact on GDP per person is much smaller and its direction unclear. In our higher migration scenario, borrowing is £19.9 billion lower by the forecast horizon, and underlying debt 3.1 per cent lower as a share of GDP. Our downside scenario is symmetric, with borrowing up by £19.9 billion and underlying debt up by 3.1 per cent of GDP.
• In our November 2023 Economic and fiscal outlook we estimated that ½ per cent higher or lower annual productivity growth would reduce or raise borrowing by more than £40 billion by 2028-29. Productivity growth across the forecast is little changed from November, but the starting level is ½ a per cent lower as revised data shows a larger population producing a similar amount of output.
• The level of economic inactivity remains significantly higher than before the pandemic. Our July 2023 Fiscal risks and sustainability report considered upside and downside scenarios for future inactivity levels. The downside scenario assumed the working-age participation rate fell 1.2 percentage points by 2027-28, leading to a 1½ per cent fall in GDP, £21.3 billion higher borrowing, and a 3.4 percentage point rise in debt as a share of GDP. The upside scenario was broadly symmetric, but with slightly smaller decreases in borrowing and debt.

1.29 There are also significant risks attached to the implementation of the Government’s stated fiscal policies. Over the medium term, these risks include:

• The fiscal forecast is conditioned on the tax-to-GDP ratio increasing by a further 1.1 percentage points from its current level to close to a post-war high of 37.1 per cent of GDP by 2028-29. Around two thirds of that increase comes from freezes to personal tax allowances which deliver around £7 billion of extra revenue for each additional year of the freeze. It also relies on the seldom-implemented indexation of fuel duty which delivers around £4.8 billion in additional revenue by 2028-29.
• The lack of detailed department-by-department spending plans beyond the current Spending Review period, which ends in 2024-25. Departmental spending over the subsequent four years of our forecast follows two overall assumptions set by the Government. These now imply no real growth in departmental spending per person over the next five years. Within this envelope the Government has committed to, among other things, an NHS workforce plan that implies real growth of 3.6 per cent each year and holding defence spending constant at 2 per cent of GDP with ambitions to raise it to 2½ per cent of GDP. Meeting these and other commitments on schools, childcare, and overseas aid spending would imply a real cut in all other departments’ budgets of 2.3 per cent a year from 2025-26. In the run-up to the Spending Reviews in November 2015 and October 2021, governments topped up the current departmental expenditure envelope by an average of £39 billion and £32 billion a year respectively.

Chapter 2: Economic outlook

Introduction

2.1 This chapter describes our latest economy forecast, summarised in Table 2.1, including:

• our conditioning assumptions, including those related to commodity prices, monetary and fiscal policy, and the global economy and exchange rate (from paragraph 2.2);
• prospects for inflation (from paragraph 2.9);
• our revised forecast for potential output, taking account of the latest population and labour force data (from paragraph 2.14);
• prospects for real GDP and the output gap (from paragraph 2.22);
• the outlook for the labour market including the projected path of employment, unemployment, and earnings (from paragraph 2.31);
• our forecasts for household incomes, the saving rate and the housing market (from paragraph 2.39);
• the outlook for the current account and nominal GDP (from paragraph 2.45); and
• how our forecast compares to recent external forecasts (from paragraph 2.48).

Table 2.1: Key economy forecast assumptions and judgements

Key forecast assumptions

Monetary policy and gilts

2.2 The expected path for Bank Rate is significantly lower than in our November forecast (left panel, Chart 2.1). When we took market determinants in the 10 working days to 23 January, market participants expected Bank Rate to fall more steeply this year to reach 4.2 per cent in the final quarter of 2024, 0.8 percentage points lower than we forecast in November. Bank Rate then reaches a lower level of 3.3 per cent at the forecast horizon, 0.7 percentage points than in our last forecast. The shallower near-term profile for Bank Rate is a counterpart to lower-than-expected inflation outturns. But expectations remain volatile, as shown by expectations for Bank Rate in 2028 oscillating between 2.7 and 4.2 per cent since our November forecast.

2.3 Gilt yields are also lower across all maturities relative to November (right panel, Chart 2.1). When we closed our market window, the 10-year gilt yield stood at 3.9 per cent, 0.6 percentage points lower than in November. Similarly, 25-year gilt yields have fallen to 4.6 per cent, 0.4 percentage points lower than our November forecast. Despite these recent falls, 10-year gilt yields are still four times higher than their level of 0.9 per cent before the invasion of Ukraine.^[1]
Partly driven by changes in Bank Rate expectations, gilt yields have also remained volatile since our November forecast, with 10-year gilt yields ranging from 3.5 per cent to 4.7 per cent.

Chart 2.1: Bank Rate and gilt yields by years to maturity

Commodity prices

2.4 Since our November forecast, market expectations for natural gas prices have continued to fall towards levels last seen before Russia’s invasion of Ukraine (left panel, Chart 2.2). Our forecast is conditioned on wholesale gas prices averaging 76 pence a therm in 2024 and 84 pence a therm in 2025. These figures are 46 and 32 pence lower, respectively, than we assumed in November. In the medium term, wholesale natural gas prices are now assumed to settle at around 78 pence a therm by 2028. This is 20 pence lower than in November, but still 28 pence higher than the pre-pandemic average of 50 pence a therm.

2.5 Market expectations for Brent crude oil prices in 2024 have also fallen relative to our November forecast, down 7 per cent to $77 a barrel (right panel, Chart 2.2). In sterling terms, oil prices in 2024 have fallen by 10 per cent to £61 a barrel reflecting the strengthening of the pound since our November forecast. In recent months, the price of crude oil has seen relatively low levels of volatility. Market expectations for 2024 prices have ranged from $74 to $86 a barrel, despite the conflict in the Middle East posing a significant risk to oil and gas markets. Box 2.2 explores the potential impact of a widening of the conflict in the Middle East on energy, goods, and economy-wide prices in the UK.

Chart 2.2: Gas and oil prices

World economy and the exchange rate

2.6 Global economic conditions have improved since our November forecast, as growth has proved more resilient to higher energy prices and interest rates than previously expected. In line with the IMF’s January World Economic Outlook Update, we expect slightly stronger global GDP growth of 3.1 per cent in 2024, with medium-term growth unchanged at around 3 per cent (left panel, Chart 2.3). Conflict in the Middle East has emerged as a major risk to the global outlook, particularly if disruption to major shipping routes fuels further inflation (see Box 2.2). The trade weighted sterling effective exchange rate has strengthened by around 2 per cent since our November 2023 forecast (right panel, Chart 2.3). As of the 10 working days to 23 January, the pound had strengthened 4 per cent against the dollar, 1 per cent against the euro and 3 per cent against the yen.

Chart 2.3: World GDP growth and the exchange rate

Fiscal policy

2.7 The large-scale fiscal support to households, firms, and public services provided during the pandemic and subsequent energy crisis has now mostly been withdrawn as both shocks have receded. Borrowing continues to fall over the forecast period as the remaining elements of crisis-related support are withdrawn, net tax continues to rise (largely through ongoing freezes to personal tax thresholds), and departmental spending falls as a share of GDP. Together, these reduce headline borrowing by 3 per cent of GDP over our forecast (with the reduction largest in 2024-25), which, on our central forecast, by 2028-29 puts the underlying debt-to-GDP ratio on a declining path. The support to demand from fiscal policy wanes over our forecast, although the combined fiscal and monetary policy mix brings output to potential and inflation to 2 per cent by our forecast horizon.

2.8 Relative to November, the path of headline borrowing is little changed across our forecast. Lower debt interest spending is offset by a modestly higher primary deficit, particularly in the final year of this Parliament. Around half of this reflects the discretionary fiscal loosening in the Spring Budget, which itself increases the primary deficit by 0.2 per cent of GDP on average over the forecast. As described in Box 2.1, we estimate the policy measures in this Budget to provide a small, temporary boost (0.2 per cent of GDP) to demand in the near term, and a similar, but permanent, boost (0.2 per cent of GDP) to supply over the medium term. The measures slightly lower near-term inflation and then raise it very slightly.

Box 2.1: The economic effects of policy measures

Our economy forecast accounts for the economic impacts of the latest announced government policies. This includes the demand-side impacts of the package as a whole, calculated using a set of fiscal ‘multipliers’ which are drawn from the empirical literature and reviewed periodically.^a These capture wider effects of fiscal policy measures on output over and above their direct effects on government expenditure, through changes to private incomes and spending. The impact of policies on the supply side of the economy is also accounted for if credible evidence suggests that measures will have a material, additional, and durable impact on potential output.

Policies announced in this Spring Budget are estimated to add £14 billion to borrowing in 2024-25, and an average of £8 billion a year from 2025-26. We estimate that these policies raise GDP by an average of 0.3 per cent across the forecast period through their impacts on both the demand and supply sides of the economy (Chart A). The near-term stimulus to demand (blue bars) almost entirely reflects the further permanent cuts to NICs and a temporary fuel duty freeze boosting post-tax real incomes. Two Budget measures are expected to have lasting supply-side effects. The additional cuts to NICs (yellow bars) and changes to the high income child benefit charge (HICBC) (green bars) are estimated to raise GDP by a total 0.1 per cent in 2024-25. The impact rises to 0.2 per cent from 2025-26 onwards as more people adjust their labour supply in response to changes in financial incentives. Box 3.2 sets out a summary of our labour supply modelling.

Chart A: Impact of policy measures on real GDP

These two Budget measures are estimated to increase labour supply by around 100,000 (in full-time equivalent (FTE) terms) in the final year of the forecast. This is additional to the roughly 200,000 estimated combined final-year boosts from the NICs, childcare, welfare reform and other measures announced at the last two fiscal events. All of these estimates are highly uncertain and subject to revision. And, as discussed in Box 3.2, the net effect of recent fiscal policy changes on labour supply also needs to take account of the negative effect of the ongoing freeze to personal tax thresholds. As we also discuss in Box 3.2, recent upward revisions to the size of the population have increased our estimates for both the cost and labour supply impact of the Autumn 2023 NICs cut (those revisions are also reflected in the impacts of frozen thresholds). Non-tax measures, in the areas of childcare and welfare reforms, are also subject to delivery risk. The expansion of free childcare from April, which we estimated would boost employment by 60,000, has received additional funding to support delivery.

We expect this Budget’s measures to reduce CPI inflation by 0.2 percentage points in 2024-25, almost entirely reflecting the effect of maintaining the 5p cut to fuel duty for one more year and freezing alcohol duty. Part of this effect unwinds in 2025-26, when the planned increase in fuel duty adds 0.1 percentage points to CPI inflation. The fiscal stimulus raises inflation by a small amount – a cumulative 0.1 percentage points by 2028-29 – which leaves the overall level of CPI broadly unchanged by the end of our forecast horizon. The cut in capital gains tax for properties increases property transactions by around 2 per cent in the near term, before tapering away over the remainder of our forecast.

a) See our 2019 Forecast evaluation report and November 2020 Economic and fiscal outlook.

b) A recent online article also set out the mechanics of our labour supply model, and how we applied it to the Autumn 2023 NICs cuts: OBR, The labour supply effects of the Autumn 2023 National Insurance Contributions cut, 2024. We also published the workings and recent application of our cost of capital model: OBR, The economic effects of full expensing, 2024.

Inflation

2.9 Inflation has fallen more sharply than we expected in November. CPI inflation dropped from a 41-year high of 11.1 per cent in October 2022 to 4.2 per cent in the final quarter of 2023, 0.6 percentage points lower than we expected in November. The surprise fall in November monthly inflation represented the largest undershoot against the industry consensus since at least 2014.^[2]
Gas and electricity bills, food, and fuel were broadly in line with our expectations in the final quarter of 2023. The downside surprise was split between lower inflation in other tradable and non-tradable goods and services. The easing in domestic inflationary pressures relative to our November forecast is consistent with our revised view that the economy is operating with slightly more spare capacity than we previously assumed (see paragraph 2.26).

2.10 We now expect inflation to fall faster than in our November forecast over the coming two years. Our central forecast sees quarterly inflation falling to the 2 per cent target in the second quarter of 2024, around one year sooner than we expected in November. We expect inflation to average 2.2 per cent over 2024, then slow to 1.5 per cent in 2025 before rising to the target rate of 2 per cent in 2028. But there is considerable uncertainty around this forecast from both domestic and external inflationary pressures, especially around energy prices. To quantify this uncertainty, we construct a fan chart that illustrates the range of possible inflation outcomes if past forecast errors were a reasonable guide to future ones (Chart 2.4). For 2025 as a whole, it suggests that there is roughly a 1 in 10 chance that average inflation will exceed 4 per cent and a roughly 1 in 10 chance that average inflation will fall below -1 per cent.

Chart 2.4: CPI inflation

2.11 The reduction in our forecast for inflation in 2024 from 3.6 to 2.2 per cent is mainly driven by significantly lower market expectations for energy prices since November. Our forecast is conditioned on the Ofgem price cap falling to £1,620 in April and averaging £1,550 between April and the end of the year, around 20 per cent lower than in our November forecast.^[3]
The direct effect of lower energy prices reduces CPI inflation by slightly under 1 percentage point on average this year and explains around two thirds of our lower inflation forecast in 2024. From late 2024 onwards, we expect the indirect effects of lower energy prices to continue to put moderate downward pressure on CPI inflation. This more than offsets a modest upward adjustment in 2024 and 2025 from the impact of the current Red Sea shipping disruptions, peaking at 0.2 percentage points in the first half of 2025. We explore the inflationary and other economic implications of a widening of the conflict in the Middle East in Box 2.2. We expect domestically generated inflation to be slightly lower over 2024 and 2025, but slightly higher in 2026 and 2027, in line with our revised output gap forecast.

Chart 2.5: Contributions to CPI inflation

Box 2.2: Economic implications of further instability in the Middle East

The prospect of a widening of the current conflict in the Middle East poses a significant risk to the global economy. So far, the economic impact has primarily been from disruption to Red Sea shipping. Shipping costs from China, measured by the Shanghai Containerized Freight Index, have risen to over twice the historical average, but remain less than half their pandemic peak. The impact on global energy markets appears modest to date. But given the region’s importance to global oil and gas supply, a surge in energy prices is a key risk in a more adverse scenario.

To explore the potential impact of further instability in the region, we use the World Bank’s October 2023 ‘large scale disruption’ scenario. This involves a cut to Middle East energy supplies comparable to the 1973 oil embargo.^a In this scenario, wholesale oil and gas prices both rise to 75 per cent above our central forecast. Gas prices peak in early 2025 and oil prices peak in mid-2024 at an all-time nominal high. After a year, prices of both commodities quickly fall back to our central forecast. We also assume that disruption to global supply chains for goods intensifies and peaks in the final quarter of 2024.^b

In this scenario, CPI inflation reaches a peak of 7.4 per cent in the second quarter of 2025 (Chart B). Informed by the recent energy shock and our analysis of its pass through to inflation, we assume inflation falls back towards baseline more slowly than it rose.^c By the end of the forecast period, the price level is 6 per cent higher than baseline. The impact is driven by:

• Direct effect of energy prices (yellow bars). Higher oil prices feed quickly into consumer prices for fuel and raise the cost of oil intensive goods such as food and transport. The Ofgem price cap slows the impact of gas prices on consumer utilities prices, which peak in mid-2025.
• Indirect effect of energy prices (green bars). The second-round effects of the energy shock then dominate. Higher energy prices pass through into tradables and non-tradables prices, with a peak effect in mid-2025, keeping inflation above baseline until the beginning of 2026.
• Impact of supply chain disruptions (purple bars). To estimate the inflation impact of further disruption to global goods trade, we compute an impulse response based on IMF research, adapted to the UK economy.^d The impact peaks in mid-2025 then fades over the next two years.

Chart B: CPI inflation in the adverse scenario

Based on a modified version of our small model,^e we assume Bank Rate would rise quickly in response to the shock, and that higher inflation and interest rates lead to a recession lasting just over a year, starting in the first half of 2025. The output gap troughs at around -5 per cent in mid-2026. Growth under this scenario quickly rebounds as inflation eases and Bank Rate falls below our central forecast. But by 2028, the economy remains around 1½ per cent smaller as some spare capacity remains. Over a longer period, we would expect the output gap to close and real GDP and Bank Rate to return to our central forecast as the effects of the shock fade. We explore the fiscal implications of this inflation scenario in Chapter 5.

As this scenario would entail a second large upside shock to inflation in quick succession, there is a risk that households and businesses begin to expect higher medium-term inflation. Should that happen, it would push domestically-driven inflation even higher, potentially prompting a stronger Bank Rate response and weaker GDP growth.

a) World Bank, Potential Near-Term Implications of the Conflict in the Middle East for Commodity Markets, October 2023.

b) As measured by the Federal Reserve Bank of New York’s Global Supply Chain Pressure Index. The index reaches the same level as at the height of the pandemic, before falling back to the historical norm over the following year and a half.

c) For more information see Box 2.1 of our 2023 Forecast evaluation report.

d) Andriantomanga, Z., et al., Global Supply Chain Disruptions: Challenges for Inflation and Monetary Policy in Sub-Saharan Africa, February 2023.

e) Given recent demand responses to changes in monetary policy, we have modified the small model so that the output gap responds to changes in the nominal Bank Rate. For more information see Working paper No.4: A small model of the UK economy, July 2012.

2.12 RPI inflation has also fallen more sharply than we expected in our November forecast. It fell from a peak of 13.9 per cent in the final quarter of 2022 to 5.5 per cent in the final quarter of 2023. This was 1.2 percentage points lower than we expected in November. In our central forecast, RPI inflation falls to 1.7 per cent by the first quarter of 2025, as CPI inflation slows and growth in mortgage interest payments eases. We then expect RPI to increase steadily in line with CPI inflation, averaging 2.0 per cent in 2025, 2.5 per cent in 2026 and 2.9 from 2027 onwards.

2.13 Annual growth in the GDP deflator, which measures the price of all domestically produced goods and services, was 7.5 per cent in the third quarter of 2023. This was 0.3 percentage points higher than we expected in November. The difference is largely explained by faster growth in the price of UK exports relative to imports. As with CPI and RPI, we expect GDP deflator growth to slow in the near term. We forecast 1.5 per cent growth in 2024 and 1.2 per cent in 2025, both around ½ a percentage point lower than our November forecast. GDP deflator growth then rises broadly in line with CPI inflation, leaving the level of the GDP deflator 0.4 per cent lower than in our November forecast in 2028-29.

Potential output

2.14 Our overall potential output forecast is little changed since November, but there have been significant revisions to its components. Chart 2.6 summarises the changes, and we describe each component in more detail in the rest of this section:

• The level of potential output in 2023 is 0.1 per cent lower than in our November forecast. This is because we judge that around half the weakness in GDP outturns since November reflects structural factors. In terms of components, while the reweighted Labour Force Survey (LFS) indicates a larger adult population, this is offset by lower trend labour force participation and average hours worked. The latter are due to the updated demographic profile in the LFS and our analysis of the effects to date of fiscal drag (see Box 3.2). Overall, our estimate for potential total hours worked in 2023 is 0.4 per cent higher. With potential output slightly lower, the level of trend productivity in 2023 is 0.5 per cent lower.
• Potential output growth from 2024 to 2028 is unchanged from our November forecast on a pre-Budget measures basis, reflecting offsetting factors. Higher net migration over the forecast raises population growth. But this is offset by a lower trend participation rate and average hours worked in the UK adult population. The latter changes are driven by updated analysis of the demographic composition of the working-age population (see paragraph 2.18). The new data show the UK population is more concentrated in lower-participation groups (under-25s and women). Ongoing effects of fiscal drag also cut potential average hours.
• Measures at this Budget raise potential output growth by a cumulative 0.2 percentage points over the forecast. As outlined in Box 2.1 this is largely from the impact of additional cuts to NICs, with changes to the high income child benefit charge making a small additional contribution.
• Taken together, these changes leave the level of potential output in 2028 marginally higher, up 0.1 per cent from November.

Chart 2.6: Potential output growth and level changes since November

Labour supply

2.15 Our forecast for potential labour supply incorporates findings from the reweighted LFS published in February. We have used this data as a guide to the UK’s current population and demographics. But there remains significant uncertainty around recent labour market data given the ONS will switch to its Transformed LFS survey later this year. This may have further implications for our understanding of labour supply and potential output.

2.16 By the forecast horizon, we expect trend total hours worked to be 0.7 per cent higher than in November (Chart 2.7). This reflects a slightly higher starting point in the latest LFS outturn, followed by slightly higher forecast growth compared to November. On a pre-Budget measures basis, growth in trend total hours worked is broadly unchanged due to offsetting factors. The adult population grows more quickly across the forecast due to our updated migration assumptions. However, the latest data on the demographic composition of the working population and updated analysis of the impact of fiscal drag on work incentives mean we expect the participation rate and average hours to fall more quickly than in November. Budget measures offset this and raise total hours worked by 0.3 per cent by the forecast horizon (Box 2.1). Assumed future levels of net migration are a key source of uncertainty around our forecast for total hours worked, and the swathe in Chart 2.7 shows how hours worked would change under our alternative migration scenarios (Box 2.3).

Chart 2.7: Trend total hours worked

Adult population

2.17 Our forecast for the adult population of 57.3 million at the end of the forecast is a million people (1.8 per cent) higher than in November, mainly driven by higher net migration. Reweighted LFS outturn data now incorporate the historically high levels of migration from the last two years and the detailed results of the 2021 Census. The UK adult population in 2023 is now estimated to have been 54.8 million, almost 700,000 people larger (1.3 per cent) than the data suggested in November. We have also incorporated a higher future path for net migration, informed by recent data outturns and updated ONS population projections published in January. As a result, net migration falls from its mid-2023 level of 670,000 to settle at the ONS’s latest medium-term estimate of 315,000 at the forecast horizon, 70,000 higher than the 245,000 assumed in November. Box 2.3 discusses our latest forecast for net migration, and its potential economic impact, in more detail.

Box 2.3: Net migration forecast and its impact on the economy

Updated ONS data published shortly after our November 2023 forecast showed large upward revisions to net migration – now estimated to have peaked at 745,000 in 2022, 140,000 higher than the initial estimate. The latest Home Office data show that visas granted (excluding visitors and temporary visas) peaked at around 1.4 million in the year to the first quarter of 2023, then fell back somewhat over the rest of the year (right panel, Chart C). Within this, work and especially dependant visas continued to grow in 2023, while student visas came down slightly and humanitarian visas fell sharply. New migrants come largely from non-EU countries, with EU net migration having fallen from over 250,000 in 2016 to turn negative in 2021 (left panel, Chart C). The large numbers and varying dynamics for types of visas create uncertainty for the path of net migration, whose potential effects we quantify in our scenarios below.

Chart C: Net migration by nationality and visas granted by type

In January, the ONS published updated net migration projections which assume net inflows follow a linear path downward from the latest outturn of around 670,000 in the year to mid-2023 to a medium-term level of 315,000 from 2027-28. This is 70,000 higher than the steady state level of 245,000 assumed in their last set of projections and used in our November 2023 forecast. The ONS based this new assumption on a 10-year historical average to mid-2023, with the choice of horizon reflecting consultation with an expert panel.^a

We use the new ONS medium-term net migration projection as the end point of our five-year forecast. While there remains considerable uncertainty about this projection, we judge that it is a reasonable estimate of the steady state level of net migration informed by consultation with migration experts. However, we have adjusted the near-term path of net inflows to incorporate a sharper short-term decline than the ONS straight-line path (Chart D). This reflects two factors:

• First, the recent rise in immigration, especially among students, should lead to higher emigration in the short term.^b
• Second, immigration policy changes announced shortly after our November 2023 EFO will likely significantly reduce inflows in the near term.^c The Home Office estimated those policies, together with those previously announced in May 2023, would cut immigration by 300,000 (mostly among dependants) under a hypothetical scenario where they were applied in the year to September 2023.^d Most of the policies will have been introduced by April 2024, so we expect significantly reduced immigration levels from that point.

This means we now expect net migration to average around 350,000 a year over the forecast period, up from 290,000 in our November forecast. Cumulatively, this adds a further 350,000 people (around 300,000 adults) to the UK population over the next five years. However, these estimates are highly sensitive to alternative assumptions for future immigration levels, the share of previous immigrants that stay in the UK and the impact of government policy.

Chart D: Net migration forecast and scenarios

In constructing our economic forecast, we estimate that migrants coming to the UK over our forecast period have a slightly higher average participation rate than the adult UK resident population over our five-year forecast. This reflects a combination of three factors:

• Age: Migrants are more likely to be of working age (16-64) when they arrive.
• Composition: The post-Brexit composition of migrants has shifted more towards groups that have historically had lower participation rates (e.g. students and dependants). However, those groups have seen rising participation rates under the new migration system. For example, LFS data suggest that migrants coming to the UK to study are now more likely to be economically active, with their participation rate increasing from 30 per cent in 2019 to 48 per cent in the year to June 2023.^e This is likely the result of the graduate route making it easier for students to enter the labour market after their studies.
• Dynamics: In the medium term, the longer migrants stay, the more likely they are to participate. While this tends to reverse after 10 years, that is beyond our forecast horizon.

Given the uncertainty around future migration flows, we explore two alternative scenarios where net migration is 200,000 higher or lower in the medium term and the total population in 2028-29 is one million higher or lower than in our central forecast (Chart D).^f In each scenario, we explore the implications of alternative assumptions for migrants’ productivity and participation in the labour force.

In the high migration scenario, net migration falls by around 160,000 from its current level to settle at 515,000 a year in 2028-29. Looking at its potential economic implications:

• The blue bars assume the additional migrants have the same productivity per hour worked as the existing population.^g This could reflect either investment responding so capital per worker is unchanged, or lower capital per worker offset by higher TFP growth. We also assume additional migrants have the same participation rate as the existing population. Based on these assumptions, GDP is 1.5 per cent higher in 2028-29, but GDP per person is largely unchanged.
• The yellow triangles show what would happen if the capital stock did not adjust at all in response to higher migration, making less capital available per worker and reducing overall productivity. GDP in 2028-29 is still higher but by a smaller amount (1 per cent), but GDP per person is then 0.4 per cent below our central forecast.
• The green diamonds show the impact of assuming additional migrants have a much higher participation rate than the adult UK population (around 90 rather than 63 per cent), consistent with a scenario where almost all adults arrive on work visas. This boosts the impact on GDP to 2.2 per cent in 2028-29 and raises GDP per person 0.8 per cent above our forecast.

The impacts in the lower migration scenario, in which net migration settles at 115,000 in 2028-29, are symmetric. These scenarios show that while additional (fewer) migrants generally boost (reduce) the level of aggregate output in the economy, the size of this impact and the effect on per person living standards is highly uncertain. The age, education, skill level, and participation rate of migrants, alongside the investment response of businesses, are all factors that determine the impact of migration on per person output. We explore the fiscal implications of net migration in Chapter 4.

Chart E: Real GDP and real GDP per person differences from our central forecast

a) The panel meeting took place on 12 December 2023, after the Government’s latest migration policies were announced.

b) This is consistent with recent academic literature, as in: Hall, T., A. Manning and M. Sumption, Projecting UK net migration, Centre for Economic Performance, Occasional Paper No.60, October 2023.

c) These policies included: restricting social care workers from bringing dependants; requiring social care firms to be Care Quality Commission (CQC) registered to sponsor visas; raising the salary threshold for skilled worker visas; changes to the shortage occupation list; and raising the minimum income required for family visas. These were in addition to previously announced policies that restricted the majority of students from bringing dependants.

d) Home Office, Legal migration statement: estimated immigration impacts, 21 December 2023.

e) Based on LFS data shown in Box 2.3 in our November 2023 Economic and fiscal outlook.

f) In line with the ONS projections, we have assumed around 85 per cent of those arriving are aged 16 or older.

g) See Migration Advisory Committee, A points-based system and salary thresholds for immigration, January 2020, which concluded “the evidence base on the impact of migration on productivity is not as strong as it is for other areas of impact… the MAC view is that there is no strong evidence for there being large effects”.

Trend participation in the labour force

2.18 We have lowered our estimates of the historical and forecast trend participation rate relative to November. It is now broadly flat at 62.8 per cent over the forecast period and, by 2028, 0.5 percentage points lower than in November. This reflects the interplay of several factors:

• The changing demographic composition of the population weighs on participation. The recent reweighting of the LFS increased the share of the 16- to-24-year-olds and women in the population. Also, ONS projections suggest that the share of the population aged 16-24 and over 65 is set to rise over the forecast period. As those groups have lower-than-average participation rates, this drags down on the overall participation rate both over history and over the forecast.
• We also expect rising health-related inactivity to continue weighing on the participation rate of prime-age adults, as well as contributing to the rise in incapacity benefit caseloads (see Chapter 4). Reweighted LFS data show the long-term sick accounted for around 30 per cent of the inactive population, on average 0.4 percentage points more than previously estimated, or around 130,000 more people.
• Previously announced measures over the fiscal events in 2023 (including the NICs, welfare and childcare reform measures) boost participation at the forecast horizon by around 0.6 per cent (for more information, see Box 3.2). Policies announced in this Budget boost participation by 0.1 per cent. Taken together, these outweigh the 0.2 per cent effect of fiscal drag.
• Net migration also raises the participation rate due to the age profile of new migrants. Recent immigration has seen a high concentration of students and dependants, who are less likely to participate than the overall population. But this is outweighed in our forecast by the age composition of migrants skewing towards prime working age, along with evidence that in the near term the longer migrants stay, the more likely they are to work.

Chart 2.8: Trend participation rate

Average hours worked

2.19 Trend average hours worked also falls over our forecast due to the same demographic and policy effects that drive participation lower. Relative to our November forecast, trend average hours start 0.4 per cent lower in 2023 due to the updated demographics of the adult population in the reweighted LFS. On a pre-Budget measures basis, growth in average hours worked over the forecast is also slightly weaker (0.1 percentage points cumulatively) due to the impact of fiscal drag on work incentives. However, this is offset by the 0.2 percentage points boost from measures in this Budget (Box 2.1). As a result, trend average hours end the forecast 0.3 per cent lower than we expected in November.

Trend productivity

2.20 With the latest data indicating that the working population was larger, but output was slightly lower, our estimate of the starting level of trend productivity in 2023 is 0.5 per cent lower than in November. However, our forecast for average trend productivity growth over 2024 to 2028 is largely unchanged at 0.9 per cent. Productivity growth is composed of capital deepening – proxied by the change in the capital stock per hour worked – and total factor productivity (TFP) – the economy’s efficiency at combining capital and labour to produce output:

• Capital deepening contributes 0.2 percentage points to average annual productivity growth over 2024 to 2028, broadly unchanged from November.
• TFP contributes 0.7 percentage points to average productivity growth over 2024 to 2028, also unchanged from November. There are two small offsetting factors. We estimate that recent TFP growth has been weaker than expected, but lower medium-term gas prices add marginally to forecast TFP growth.

2.21 The net effect of these changes leave the level of trend productivity 0.6 per cent lower in 2028 than in November, mostly reflecting recent data showing more hours worked within a largely unchanged level of output.

Real GDP and the output gap

Real GDP

2.22 Real GDP grew by 0.1 per cent in 2023, 0.4 percentage points weaker than anticipated in our November 2023 forecast. This reflects ONS revisions, which lowered growth in the third quarter of 2023, and weaker-than-expected GDP growth in the fourth quarter (released after we finalised our forecast). Timely survey indicators generally suggest economic activity has stabilised and may be picking up slightly. For example, the S&P Global/CIPS UK composite PMI points to an expansion in activity in recent months, and the flash estimate for February edged up to around the historical average.

2.23 We therefore forecast GDP growth to pick up only slightly to 0.8 per cent in 2024, as the economy continues to be constrained by weak real wage growth, the ongoing effects of recent interest rate rises and fading fiscal support. We expect growth to accelerate to around 2.0 per cent in 2026 as interest rates decline and spare capacity in the economy is used up. The output gap closes at the end of 2027 and growth then returns towards our estimate of potential, averaging 1.8 per cent in 2027 and 2028. At the end of the forecast period, the level of GDP is 0.1 per cent higher than in November.

2.24 Risks to our real GDP forecast remain elevated. As always, the outlook for productivity growth is our most important and uncertain forecast judgement. The effects of subdued investment, the energy price shock and Brexit compound the ongoing weakness seen since the financial crisis. There are also significant risks to our participation and average hours forecasts from the effects of an ageing population and the outlook for long-term sickness among those of working age. As demonstrated by the scenarios in Box 2.3, there is also significant uncertainty around the level and composition of net migration over the forecast. The financial crisis, pandemic and energy price shock also showed that there is always the risk of an unforeseen external shock having a large impact on the UK. If past forecast errors are a reasonable guide to future ones, then the likelihood that annual GDP growth will be negative in 2024 is 2 in 5.

Chart 2.9: Real GDP growth forecast

2.25 Real GDP per person has fallen continuously since the first quarter of 2022 (the longest sustained decline since records began in 1955) and fell by 0.7 per cent in 2023. In our central forecast, it troughs 1¼ per cent below its pre-pandemic peak in 2024 and only recovers to that peak in 2025. The level of real GDP per person is expected to be ¾ a per cent lower in 2028 than in our November forecast. This is driven by downward revisions to the starting point of GDP per person by the ONS, due to higher population outturns, and our lower forecast for trend participation. Our migration scenarios also demonstrate that while higher net migration has been a key driver of GDP growth in recent years, under plausible assumptions, it may not have a substantial impact on output per person.

Chart 2.10: Level of real GDP and real GDP per person

Output gap

2.26 We judge that the economy has been operating with slightly more slack than we expected in November. This is consistent with the faster-than-expected drop in domestically generated inflation and widening survey indicators of spare capacity. We now estimate that the output gap was -0.1 per cent in the third quarter of 2023, 0.2 percentage points lower than in November. But the directly unobservable nature of the output gap means estimates of its size are extremely uncertain. We therefore use several models to inform our estimate (Chart 2.11). Most of our models suggest the output gap was around zero in the second half of 2023, with business surveys suggesting there is still some excess demand in the economy.

2.27 The output gap is now forecast to widen from close to zero to trough around -1 per cent at the end of 2024. Policy measures in the Budget that boost demand are expected to lift the trough in the output gap by 0.1 percentage points. As GDP growth in our forecast picks up above potential in 2025, the negative output gap closes in the second half of 2027. This is around half a year earlier than estimated in November, largely due to lower market expectations for interest rates.

Chart 2.11: Output gap

Expenditure composition of GDP growth

2.28 Real consumption was weaker than forecast in the second half of 2023 and is expected to remain subdued in the near-term, growing 0.7 per cent in 2024. We forecast household spending growth to pick up from 2025, reaching an average of 2.0 per cent in 2025 and 2026, around ⅔ a percentage point higher than in November. The stronger consumption profile reflects higher household disposable incomes, a sharper slowdown in inflation, and lower interest rates. In the central forecast we assume the saving ratio will remain elevated in the near term before slowly declining to its long-run average (see paragraph 2.41). There are risks in both directions around our consumption forecast, reflecting uncertainty around future real wages, household saving and house prices.

2.29 Business investment is volatile and has historically seen large revisions but was 3.0 per cent higher at the end of 2023 than we expected in November. We expect business investment to contract in the near term as past increases in interest rates raise the cost of capital and weigh on spending. Investment then strengthens as interest rates fall and economic demand picks up. Despite stronger medium-term growth, the nearer-term weakness means that cumulative investment over 2024 to 2028, and therefore the increase in the capital stock, is largely unchanged from November.

2.30 We forecast that trade volumes will be subdued in the next few years due to sluggish growth in the UK and global economies, alongside the evolving impact of Brexit. Recent trade data have been volatile and subject to large revisions. However, they remain broadly consistent with our assumption that Brexit will reduce the UK’s trade intensity (exports plus imports as a share of GDP) by 15 per cent in the long term (see Box 2.4). From 2024 to 2028, we expect export volumes to grow by an average of 0.3 per cent, 0.1 percentage points higher than November. Import volumes are expected to rise 0.1 per cent a year on average over 2024‑2028, 0.4 percentage points higher than November, consistent with higher domestic demand. Recent disruptions in the Red Sea present a risk to global trade and energy prices (see Box 2.2), but our central forecast does not include a material impact on UK trade.

Chart 2.12: Contributions to real GDP growth

Box 2.4: How are our Brexit trade forecast assumptions performing?

Since the June 2016 EU referendum, our forecasts have assumed that the volume of UK imports and exports will both be 15 per cent lower than if we had remained in the EU. We assume that the resulting reduction in the trade intensity of GDP will lead to a 4 per cent reduction in the potential productivity of the UK economy (relative to remaining in the EU), with the full effect felt after 15 years. A decline in trade intensity plausibly lowers productivity because trade, among other channels, fosters competition and allows countries to specialise in activities where they are relatively more efficient.^a This box assesses these assumptions against the latest evidence.^b

The Trade and Cooperation Agreement (TCA), signed in December 2020, set the terms of the post-Brexit trading relationship between the UK and EU. However, the application of the TCA remains ongoing, with the UK implementing physical inspections on imports from the EU from April 2024 and further declaration requirements from October 2024.

Overall trade intensity in the UK and the rest of the G7

Trade volumes in all advanced economies declined sharply at the height of the pandemic in 2020. However, UK trade intensity (exports plus imports as a share of GDP) has not recovered in line with other G7 countries since then (Chart F). In the third quarter of 2023, UK trade intensity remained 1.7 per cent below its pre-pandemic level from 2019. By contrast, it had risen 1.7 per cent above pre-pandemic levels on average in the rest of the G7.

Chart F: Trade intensity in the UK and rest of the G7

More stringent regulation of trade flows, the expectation of further regulatory tightening, and uncertainty over future trade policy may all have weighed on trade between the UK and EU over this period. Goods trade between EU countries and goods trade between the EU and the rest of the world grew more than a third between 2019 and 2022 in current prices. Meanwhile, goods trade between the UK and EU only grew by only around 10 per cent over the same period. An ONS survey of businesses from early 2024 indicates around half of exporting firms and two thirds of importing firms have reported extra costs associated with the changes in regulation since the end of the transition period. Similarly, the Bank of England Agents’ survey suggests EU trade frictions have weighed on export demand and may compound as new UK-EU regulations come into operation this year.^c

Composition of trade in the UK and the rest of the G7

Within UK trade, there has been a significant and growing divergence between the performance of goods and services since the pandemic (Chart G).

• Growth in UK goods trade (exports plus imports) has fallen well behind the rest of the G7. At the end of 2023, UK goods trade was around 10 per cent below 2019 levels, while it was around 5 higher on average for the rest of the G7 in the third quarter.
• Meanwhile, UK services trade growth has been the strongest in the G7. It reached around 12 per cent above 2019 levels at the end of 2023, versus around 9 per cent above in the rest of the G7 in the third quarter.

The UK’s differential performance between goods and services trade post-Brexit likely reflects several factors. First, global trade in services, where the UK has a relative comparative advantage, has grown faster than global goods trades since 2008. Second, post-Brexit trade barriers have created more significant frictions for goods than services. Third, the UK is less dependent on the EU for services exports than goods exports – around a third of UK service exports go to the EU, compared to roughly half for goods.^d Fourth, digitalisation makes trade in some services easier and less dependent on physical proximity than in the past.^e

Chart G: Goods and services trade in the UK and rest of the G7

What is driving the growth in UK services trade?

UK services trade has continued to grow strongly, including with the EU, despite the increase in trade barriers post-Brexit. Looking at its sectoral composition, around two thirds of the growth in services trade volumes since 2019 has been driven by the ‘other business services’ sector, which includes management consulting, research & development, and advertising (Chart H). By contrast, exports of financial services and transport have lagged other sectors, declining 5.9 per cent and 2.0 per cent respectively. These are also sectors which are more likely to have been impacted by Brexit frictions.^f

The recent strong growth in the UK’s exports in ‘other business services’ could reflect several factors. First, trade barriers with the EU may be lower for these kinds of services than goods or other, more highly regulated, services such as banking.^g Second, there has been particularly strong growth in service exports to the US, possibly capturing US firms outsourcing work to the UK, supported by the recent weakness in the pound.^h Finally, there is some evidence that services firms may have circumvented trade barriers by selling through foreign affiliates.ⁱ

Chart H: Breakdown of growth in UK service exports since 2016

Overall, our assumptions about the impact of Brexit appear to be broadly on track and recently published studies are also broadly consistent with these estimates.^j However, it remains hard to draw firm conclusions given the challenges of disentangling the simultaneous impacts of Brexit, the pandemic, and other geopolitical developments affecting UK and global trade. Trade data are also volatile and prone to revision, particularly trade in services. Moreover, the full implementation of the TCA will further increase barriers to trade in goods with the EU. We expect the total impact of Brexit to be realised several years after full implementation of these barriers. In the meantime, we will keep our Brexit assumptions under review.

a) OBR, Brexit and the OBR’s forecasts, October 2018. See Box 2.2 for further analysis of the trade-productivity link.

b) See Box 2.4 of the March 2023 Economic and fiscal outlook for the economic impact of Brexit on investment and migration.

c) See Box D: Agents’ update on business conditions of Bank of England, Monetary policy report, February 2024.

d) Resolution Foundation, Ending stagnation: a new economic strategy for Britain, December 2023.

e) World Trade Organisation, World trade report 2019: The future of services trade, October 2019.

f) See Springford, J., Brexit, four years on: answers to two trade paradoxes, January 2024.

g) Intra-EU goods trade is around 60 per cent higher than the EU’s trade in goods with the rest of the world. Intra-EU trade is broadly in line with EU trade in services with the rest of the world. See European Commission, Single market scoreboard, Integration of goods and services, February 2024.

h) Financial Times, Britain, consultation nation, January 2024.

i) For example, a UK software provider could send its product directly to an EU client but may be subject to stricter regulation under the TCA. So instead, the UK firm may opt to provide the software through an EU affiliate to avoid these restrictions. See Breinlich, H. and M. Magli, Should we stay or should be go? Firms’ decisions on services mode of supply, forthcoming.

j) Springford, J., Brexit, four years on: answers to two trade paradoxes, January 2024 estimates the hit to UK trade from Brexit is consistent with a 4-5 per cent reduction in GDP to date. NIESR, Revisiting the effect of Brexit, November 2023 estimates GDP will be 5-6 per cent lower by 2035. Goldman Sachs, UK – The structural and cyclical costs of Brexit, February 2024 estimates a GDP impact of 5 per cent since the referendum.

Labour market

2.31 Our forecast incorporates the reweighted ONS LFS estimates, released in February, of the size and composition of the UK population. While this improves the representativeness of the data, problems with low sample sizes and volatility in recent periods will take longer to resolve.^[4] We have therefore placed less weight on the LFS data for the last six months when thinking about the near-term outlook. Instead, we look at a wider range of indicators such as administrative data sources and surveys.

Participation

2.32 Following the pandemic, the number of inactive working-age people rose by around 750,000 to its peak in mid-2022 and remained close to these levels at the end of 2023. Previous LFS estimates had suggested a fall in inactivity in early 2023. But in the new data, inactivity increased by 120,000 from the end of 2022 to the end of 2023 to return to an 11-year high of 9.3 million (left panel, Chart 2.13). Inactivity due to long-term sickness reached 2.8 million, around a third of the total (right panel, Chart 2.13) and a rise of 200,000 over the year to 2023. This picture ties in with the sharp rise in the number of incapacity benefit claimants,^[5] which rose 20.3 per cent between 2019-20 and 2022-23, from 2.5 million to 3.1 million. Inactivity for other reasons in the reweighted LFS is also different to previous estimates. The shares of students and carers in the inactive population are now respectively 0.6 and 0.3 percentage points higher, while the share of retirees is 0.5 percentage points lower.^[6]

Chart 2.13: Inactivity of 16-to-64-year-olds

2.33 The reweighted LFS data show the participation rate was 62.8 per cent in the final quarter of 2023. This is 1.5 percentage points below its peak of 64.3 per cent in the first quarter of 2020 and 0.5 percentage points below our November forecast. A combination of demographic effects and the impact of Budget policies outlined in paragraph 2.14 mean the participation rate remains broadly flat across the forecast. At 62.8 per cent in 2026, our forecast is very close to the Bank of England’s forecast of 62¾.

Unemployment and employment

2.34 From a post-pandemic trough of 3.6 per cent in the three months to August 2022, the LFS unemployment rate increased to 4.3 per cent in the three months to July 2023. According to the reweighted LFS data, unemployment has since fallen to 3.8 per cent in the fourth quarter of 2023. In contrast, claimant count data measuring the number of people on unemployment benefits have remained flat in recent months, while the redundancy rate has been on a slow upwards trend since the middle of 2022. The number of vacancies per unemployed has also fallen further over the last year, and recent surveys suggest that recruitment and retention are now less of an issue for firms.

2.35 We judge that this and wider evidence is consistent with some further loosening in labour market conditions. We expect a moderate rise in the unemployment rate, peaking at 4.5 per cent in the last quarter of 2024, in line with our forecast for subdued economic growth and increasing slack. The peak in unemployment, which represents a total of around 1.6 million people looking for work, is marginally lower than in our November forecast, though it comes half a year sooner. The unemployment rate is then forecast to decline to its estimated structural level of 4.1 per cent in 2028.

2.36 LFS reweighting has increased the level of employment, but the employment rate remains close to post-pandemic lows and employment growth has been weakening across a range of indicators. The employment rate is expected to fall from 60.2 per cent in the final quarter of 2023 to 60.0 per cent in the final quarter of 2024. By the end of the forecast, we expect the employment rate to reach 60.1 per cent, 0.5 percentage points lower than in our November forecast, reflecting a lower medium-term participation rate. This leaves the employment rate 0.6 percentage points lower than its post-pandemic peak of 60.8 per cent.

Chart 2.14: Unemployment and employment rates

Average earnings

2.37 From a thirty-year peak of close to 7 per cent in 2023, we expect nominal average pay growth to slow to 3.6 per cent in 2024. This is 0.2 percentage points lower than our November forecast, pulled down by a more rapid slowdown in inflation and some loosening in labour market conditions. There is some stickiness in pay settlement expectations, which remain between 5 and 5½ per cent for 2024. But we do not expect repeats of the strong bonus growth and one-off cost of living payments seen last year. As inflation falls below target and slack remains in the labour market, we expect nominal earnings growth to fall to around 2 per cent in 2025 and 2026. In our central forecast earnings growth rises to around 2½ per cent in the final year as the output gap closes and inflation returns to target. The level of wages and salaries, a key determinant of our fiscal forecast, is 0.3 per cent lower at the forecast horizon than in November, broadly in line with our revision to nominal GDP. Weaker nominal earnings growth in the medium term more than offsets modestly faster growth in employees.

2.38 In real terms, we expect earnings to grow 1.4 per cent in 2024, 1.1 percentage points stronger than we forecast in November, as wage settlement expectations have held up despite falling inflation. Real earnings growth then slows to 0.3 per cent in 2026 before gradually recovering closer to productivity growth of 1 per cent at the forecast horizon. A key risk to the near-term outlook is if the labour market proves tighter than expected, keeping pay settlements higher over the coming year. As well as posing an upside risk to our forecast for inflation, this is also a key sensitivity for our fiscal forecast. Higher-than-expected wage growth would, all else equal, drive stronger tax receipts.

Chart 2.15: Nominal and real earnings growth

Household income and saving

2.39 Based on ONS outturn to the third quarter of 2023, we estimate that real household disposable income (RHDI) per person, a measure of living standards, grew 0.8 per cent in the 2023-24 fiscal year. This follows a fall of 2.2 per cent in 2022-23 – the largest year-on-year drop in living standards since records began in the 1950s. And rather than continue falling in 2024-25 as we forecast in November, we now expect RHDI per person to continue its rebound and grow on average by 1 per cent a year over the forecast period. Falling inflation in 2023-24 and 2024-25 (blue bars in Chart 2.16) partly reflect lower commodity prices which reverse the terms of trade hit seen in 2022-23 and provide a boost to RHDI per person growth. While the pass-through to deposit rates of increases in interest rates supported the growth in RHDI per person in 2022-23 and 2023-24, we expect this to reverse over the course of 2024-25 as higher rates feed through into most fixed-term mortgage contracts (purple bars). Tax and benefits policy contributes to RHDI growth in 2024-25 but, over the forecast period, lower uprating of indexed welfare payments acts as a drag on growth (yellow bars).

2.40 With RHDI per person growing from 2023-24 rather than falling for three consecutive years, we now expect it to regain its pre-pandemic peak in 2025-26, two years earlier than in the November forecast. This is largely driven by a faster recovery from the terms of trade shock and a higher share of whole-economy income going to labour in the near term. At the forecast horizon, RHDI per person is around 1 per cent higher than in our November forecast. Lower consumer prices raise RHDI per person by 1.3 per cent relative to November. Consistent with lower inflation, labour incomes per person subtract 0.7 per cent by 2028-29 (the offset is not complete due to the improvement in the terms of trade). And forecast and policy changes to household benefits and taxes raise RHDI per person by 0.6 per cent. The 2 pence cut to the main rates of NICs announced in this Budget alone is expected to directly boost real household incomes by 0.5 per cent. This adds to a boost of similar size from the NICs cut announced in the Autumn Statement 2023. These estimates do not account for behavioural responses to the policy or fiscal drag arising from frozen tax thresholds in combination with strength in nominal earnings growth (described in Box 3.2).

Chart 2.16: Real household disposable income per person

2.41 We expect the adjusted household saving rate (excluding pensions) to remain above 4 per cent in the near term, before declining slightly to 3½ per cent at the forecast horizon. We estimate saving as a share of household disposable income was 4.3 per cent in 2023, 1.1 percentage points higher than expected in November, as consumption was somewhat weaker and incomes higher. In the near term, a weakening labour market should motivate higher precautionary saving, while elevated deposit rates encourage savings that yield interest returns. In the final years of the forecast, adjusted household saving is projected to return only slowly towards its long-run average as interest rates remain high. The headline saving ratio, which also accounts for pension fund accumulation, is estimated to have reached 9.1 per cent in 2023 and is forecast to fall to just below 8 per cent by 2028. This is similar to the November forecast as higher non-pension saving is offset by lower pension valuations given recent movements in interest rate expectations.

Credit conditions and the housing market

2.42 Average interest rates on the stock of mortgages are expected to rise to a peak of 4.2 per cent in 2027. This is up from a low of 2 per cent at the end of 2021 and above the average mortgage interest rate in the 2010s of around 3 per cent. But it is 0.8 percentage points lower than our November forecast, driven by substantial falls in market expectations for Bank Rate. However, there are significant risks to our mortgage rate forecast, demonstrated by the large movements in Bank Rate expectations since November. This also poses a risk to household incomes, residential housing transactions and house prices.

2.43 As Bank Rate rose to reach its current level in the third quarter of 2023, housing transactions fell to 256,000, down more than 40 per cent from the post-pandemic peak. Leading survey indicators from the Royal Institute of Chartered Surveyors suggest that the market will remain weak, but there are signs that demand is beginning to recover as interest rates on new mortgages continue to fall in anticipation of cuts in Bank Rate. Our central forecast is for residential property transactions to be broadly flat 2024, compared to a fall of 7 per cent in our November forecast. Transactions then recover in our forecast, reaching pre-pandemic levels in early 2025, two years before we expected in November.

2.44 In our central forecast, we expect house prices to fall around 2 per cent in 2024, slightly under half of the 5 per cent we expected in November. This is mainly due to our lower mortgage rate forecast. The average house price in the UK is forecast to fall to slightly under £275,000 in the final quarter of 2024. Supported by falling new mortgage rates, we then expect house prices to grow around 2 per cent in 2026, and around 3½ per cent in 2027 and 2028. That would see nominal house prices surpass their historical peak in the first quarter of 2027.

Chart 2.17: Mortgage rates and house prices

The current account and sectoral net lending

2.45 The current account deficit narrowed from 3.1 per cent of GDP in 2022 to an estimated 2.6 per cent in 2023. This was driven by a narrowing in the trade deficit due to a recovery in the UK’s terms of trade – the price of exports relative to imports. The terms of trade deteriorated in 2022 after Russia’s invasion of Ukraine prompted a surge in imported energy prices. The terms of trade have now recovered to early 2021 levels as energy prices have fallen back and the price of services exports have risen. After a surplus in the previous two years, net investment income is estimated to have returned to a deficit in 2023 as income earnt by overseas investors on UK investments rose by more than interest income earnt on the UK’s overseas investments. We expect the current account deficit to widen slightly in 2025, driven by a further deterioration in net investment income. It is then forecast to narrow gradually to around 2½ per cent of GDP in 2028, driven mostly by the trade balance.

2.46 Due to the changes in our earnings and saving rate forecasts, we expect the household sector surplus to reach around 2¾ per cent of GDP in 2025. Over the remainder of the forecast, household sector balances decline and the surplus falls to around 1½ per cent. Following a period of volatility during the pandemic and energy crisis, we forecast the corporate sector to turn from surplus to deficit in 2025. At the same time, we expect the fiscal deficit to fall over the course of the forecast.

Chart 2.18: Sectoral net lending

Nominal GDP and its composition

2.47 Our central forecast, together with small revisions in the historical data, leaves the level of nominal GDP 0.3 per cent lower in 2028-29 than we forecast in November (a difference of £11 billion). From a 0.2 per cent higher starting point in 2023-24, nominal GDP grows by 17 per cent over the next 5 years, 0.6 percentage points slower than we expected in November. Stronger cumulative real GDP growth is more than offset by slower growth in the GDP deflator. Breaking down these changes, we see that:

• On the income side, the near term the difference in nominal GDP is driven by lower corporate profits, which are only partly offset by upward revisions to labour incomes. This is driven by recent outturns and wage settlement expectations holding up relative to falling inflation, which should squeeze corporate margins. By the forecast horizon, both the labour and profit shares move towards their long-term averages as firms rebuild margins. Both lower labour incomes and corporate profits contribute around 0.1 percentage points of the difference in nominal GDP by 2028-29.
• On the expenditure side, the near-term downward revision to nominal GDP is largely driven by lower nominal consumption growth, an important driver of our receipts forecast. In 2024-25, lower real consumption and a lower consumer expenditure deflator lower nominal GDP by 0.8 per cent compared to our November forecast. By 2028-29, stronger real consumption growth partly offsets a lower expenditure deflator to leave the contribution to nominal GDP from nominal consumption around 0.1 per cent lower than in November.

Chart 2.19: Nominal GDP changes since November

Comparison with external forecasters

2.48 Our central forecast for real GDP growth from 2024 to 2028 is on average 0.3 percentage points higher a year than the average of other forecasters. This difference will partly be due to other forecasters not incorporating the impact of policy measures announced in this Budget but will additionally reflect varying assumptions around underlying growth in the economy’s productive capacity. We also expect stronger growth than the Bank of England in all years of their forecast. By its forecast horizon of 2026, cumulative GDP growth is 2.7 percentage points lower than in our central forecast. This difference is likely driven by several factors including: a projected negative output gap at the end of the Bank’s forecast period, more pessimistic assumptions surrounding productivity growth, and differences in the population and labour market data underpinning each forecast due to the timing of recent ONS publications.

Chart 2.20: Comparison of forecasts for cumulative growth in real GDP

2.49 Relative to the average of external forecasts and the Bank, our forecast for CPI inflation is lower in 2024, and materially lower in 2025 and 2026. This likely reflects a range of factors including the different vintages of energy prices underpinning our forecasts, and different assumptions for Bank Rate and spare capacity. Our unemployment projection for 2025 is 0.4 percentage points below the Bank’s forecast. The Bank also sees the unemployment rate climbing to 5 per cent in 2026 as the output gap remains negative.

Table 2.2: Comparison of GDP growth, CPI inflation and unemployment forecasts

Chapter 3: Policy measures

Introduction

3.1 This chapter:

• sets out the policy measures announced in this Budget, and in the period since the Autumn Statement in November, including how they have been incorporated in our forecast and the uncertainties around them;
• provides an update on selected previous measures; and
• discusses policy risks, which are possible measures that are yet to impact our central forecast.

Policy announcements in the March 2024 Budget

3.2 This forecast incorporates the economic and fiscal implications of all the policy measures that have been announced since the November 2023 Autumn Statement. The Chancellor has chosen to spend the improvement in the pre-measures forecast since November on Budget measures. In aggregate, we estimate these measures increase borrowing by an average of £8.8 billion a year (0.3 per cent of GDP). Box 3.1 shows this continues a historical pattern of discretionary fiscal loosening in the face of fiscal improvements.

3.3 Tax-cutting measures raise borrowing by an average of £12.6 billion a year over the forecast period (Table 3.1) including:

• A further two-percentage point cut in NICs, which is effective from 6 April and applies to employees and the self-employed, costs an estimated £10.5 billion a year. The net cost of the new policy is reduced to £8.9 billion once we include its indirect effects – the revenue impact of our judgement that the measure increases total hours worked by 0.3 per cent, or 98,000 in full-time equivalent terms. Box 3.2 presents our assessment of the labour supply effects of the major income tax and NICs measures that have been announced during the past three years.
• The cancellation of the planned rise in fuel duty, at an estimated cost of £3.1 billion in 2024-25 and an annual average of £0.8 billion thereafter, and a delay to the planned uprating in alcohol duties, which costs an additional £0.4 billion a year from 2025-26 onwards, while other smaller tax cuts amount to £0.4 billion a year.

3.4 The cost of these tax cuts is partially recouped in the medium-term by a set of revenue-raising measures, the most significant of which is the reform of the current non-domicile regime, which we estimate raises £2.6 billion by 2028-29. The estimated cost of other tax-raising measures include:

• HMRC compliance measures, which increase receipts by £1.1 billion a year;
• a one-year extension to the energy profits levy, which raises £1.2 billion in 2028-29;
• two new taxes – a duty on the e-liquid that is used in vaping, and the carbon border adjustment mechanism – that together generate £0.7 billion by 2028-29;
• reforms to the tax treatment of furnished holiday lettings and the abolition of multiple dwelling relief raise £0.6 billion of additional receipts by 2028-29; and
• other tax-raising measures that increase yield by £0.5 billion a year during the final three years of the forecast.

3.5 The spending decisions announced in the Budget largely net off in fiscal terms. The Government has reduced the assumed aggregate level of current departmental spending by £0.8 billion per year on average beyond 2024-25. It is also allocating an additional £0.9 billion of capital departmental spending, over the three years from 2025-26 to 2027-28, for a public sector productivity programme focused on the NHS. The decision to broaden the scope of child benefit increases child benefit spending by around £0.4 billion a year.

3.6 The indirect effects of these policies lead to a boost to demand in the near-term and a permanent supply-side increase. The former is primarily driven by the boost to household incomes provided by the near-term tax-cuts, including the continued fuel duty freeze and the cut in NICs. The latter, in conjunction with the child benefit measure, serves to boost the labour supply by improving work incentives. This is somewhat offset by the additional debt interest costs associated with financing the measures. Overall, the indirect effects of the policy package lowers borrowing by an average of £0.8 billion a year.

Table 3.1: Total effect of Government decisions since November 2023

Box 3.1: Policy responses to revisions in the pre-measures forecast

In this Budget, the Chancellor has chosen to loosen fiscal policy by 0.3 per cent of GDP per year on average over the next 5 years, in response to an 0.1 per cent of GDP improvement in the underlying (pre-measures) fiscal outlook over that same period.^a As Chart A shows, this continues the consistent pattern of expansionary fiscal policy packages that have been announced in recent years. Since December 2014, fiscal policy has been tightened in only two out of twenty fiscal events. This pattern is partially explained by the Government using fiscal policy to support the economy in the face of the series of shocks it has faced during that time. It contrasts sharply with the persistent tightening of fiscal policy between 2010 and 2014.

Chart A: Cumulative size of fiscal policy packages, 2010 to 2024

Chart B shows how policy decisions since 2010 have responded to forecast changes in the underlying fiscal position. Our forecast is conditioned on stated Government policy and so asymmetric policy responses to future shocks introduce a potential source of bias. The chart plots the aggregate size of discretionary policy packages in each fiscal event (vertical axis) against the size and direction of the underlying forecast revision (horizontal axis).^b It shows that:

• In every instance of good news on the underlying fiscal position, the policy response has been to spend at least some of the windfall (top left quadrant). Typically, less has been spent than the underlying fiscal improvement (most of the dots in the top left quadrant are below the 45-degree dashed line that signifies a one-for-one change).
• The responses to deteriorations in the underlying fiscal position have been more varied – with a mixture of fiscal tightening (bottom right quadrant), fiscal loosening (top right quadrant) and several instances of a relatively neutral fiscal package. In no case has the tightening of fiscal policy been equal to or greater than the underlying forecast deterioration (there are no dots below the 45-degree line in the bottom right quadrant).
• This Budget falls into the top left quadrant, meaning that the response to the improvement in the pre-measures fiscal position (by 0.1 per cent of GDP) since our November forecast, has been to loosen fiscal policy (by 0.3 per cent of GDP).

Chart B: Policy responses to underlying fiscal forecast revisions

a) A ‘loosening’ of fiscal policy is defined as when the cumulative 5-year fiscal impact of a policy package increases borrowing. A ‘tightening’ of fiscal policy is the reverse. This analysis excludes the cyclical component of our forecast revisions.

b) The cumulative effect of policy decisions across each forecast period are shown on the vertical axis and are measured as the total cash effect on borrowing over the five-year period, as a share of total nominal GDP over the same period. This expresses the typical ‘£ billion’ effect of a fiscal loosening or tightening in units that do not rise over time simply because the cash size of the economy is growing. Our pre-measures forecast changes are shown on the horizontal axis using the same metric. The effect of major classification changes has been stripped out.

Reductions in employee and self-employed NICs

3.7 The Chancellor has followed the cuts in the rate of National Insurance Contributions (NICs) that he announced in November (and which came into effect for employees in January) with further reductions in this Budget, once again benefiting around 28 million employees and over 2 million self-employed individuals. The two components of the policy are:

• A 2 percentage point cut in the main rate of employee Class 1 NICs, from 10 per cent to 8 per cent, effective from 6 April 2024. This follows the 12 per cent to 10 per cent cut that was announced in the Autumn Statement and implemented on 6 January. This will apply to employees with earnings greater than £12,570 on an annualised basis, with the maximum gain received at and above the upper earnings limit of £50,270. We estimate that 27.6 million employees will benefit, with annual gains of £303 for basic rate payers, £646 for higher rate taxpayers and £705 for those on the additional rate. Taking the Autumn Statement and Budget measures together, the combined gains are £606, £1,291 and £1,413 respectively.
• A 2 percentage point cut in the main rate of Class 4 NICs paid by the self-employed, from 8 per cent to 6 per cent, effective from 6 April 2024. This is on top of the 1 percentage point cut that was announced in November, which is also being implemented on 6 April (plus the removal of the requirement to pay Class 2 NICs). The current measure applies to self-employed individuals with profits greater than £12,570, with the maximum gain received at and above the upper profits limit of £50,270. We estimate it benefits 2.2 million self-employed individuals in 2024-25, with the annual average gain ranging from £235 for basic rate taxpayers, £635 for those on the higher rate, and £710 for additional rate taxpayers. For the two measures combined, these gains are £353, £946 and £1,058 respectively.

3.8 This measure is estimated to have an average five-year cost of £10.3 billion a year. This is on top of the revised £10.3 billion cost of the November measure and means that the two measures combined reverse around a half of the tax raised from the freezes in the personal tax thresholds that were announced between March 2021 and November 2022.

3.9 Table 3.2 shows the main components of the costing, which primarily reflects:

• Its estimated static cost of £11.3 billion in 2028-29 – this is the amount it would cost before we account for behavioural responses.
• A direct behavioural response to the measure, namely that it lowers the financial incentive for tax-motivated incorporations (TMIs), which we estimate will be reduced by a cumulative 83,000 TMIs by 2028-29. This generates a net increase in receipts of £0.6 billion, with additional amounts of income tax and NICs outweighing the loss in corporation tax. In cumulative terms, TMIs are expected to be 151,000 lower due to the combined effect of the two measures by 2028-29.
• Table 3.2 also reflects the indirect behavioural effects of the measure, which we estimate boosts labour supply by 0.3 per cent or 98,000 in full-time equivalent (FTE) terms (see Box 3.2 for further detail).

3.10 The dynamic behavioural effects combine to lower the cost of the NICs cut measure by an estimated £2.3 billion (around 20 per cent) relative to its static cost.

Table 3.2: Costing of the cuts in the main rates of employee and self-employed NICs

Reform of the current regime for non-domiciled individuals

3.11 The Chancellor has announced that, from April 2025, the current ‘remittance basis’ (RB) tax treatment for non-domiciled UK residents will be abolished and replaced with a new regime. Under the current RB, non-domiciled individuals have the option of paying UK tax only on their UK income and gains plus any foreign income and gains (FIG) that they choose to bring into (remit to) the UK. Taxpayers that are eligible can choose each year whether they are taxed under the RB or on the ‘arising basis’, where they are taxed on their worldwide income and gains.^[7]

3.12 The new regime is open to a smaller number of individuals, and its main characteristics are:

• As with the current regime, individuals must opt into it, but can only do so if they come to the UK after a period of at least ten years of non-residence in the UK. Eligibility is determined by residence in the UK rather than their domicile, such that the reform effectively removes the concept of a ‘non-domicile’. The costing assumes that around 10,500 individuals will be eligible for the new regime in April 2025.
• The new regime is only available for the four years after first becoming tax resident in the UK, irrespective of whether the individual is currently UK-resident.^[8] Those that opt in are exempt from paying UK tax on FIG, including the FIG remitted to the UK, for their first four years of UK tax residence. This is more generous than under the current RB, where FIG can only be remitted tax-free so long as it is invested in a qualifying UK business.
• Unlike the RB, there is no annual charge to access the new regime, but in common with the current regime, those that opt in lose their tax-free personal allowance for income tax and the annual exempt amount for capital gains tax.^[9]
• Overseas workday relief (OWR) is being retained and will be available to anyone who opts into for the new regime, for their first three years of UK tax residence. Under the existing OWR, individuals only pay UK tax on the portion of their employment income that relates to their UK workdays, effectively gaining a tax relief for work that is undertaken outside the UK.
• Ineligible individuals are subject to UK taxation on all newly arising FIG, effectively treating them as UK domiciles. The costing assumes this affects around 5,500 individuals in April 2025.

3.13 Alongside the introduction of the new regime, from April 2025 offshore protected trusts will be brought into the scope of UK taxation for all newly arising FIG. This ends the protection offered to some deemed domiciles from 2017. But this measure does provide three new avenues to protect those that are ineligible for the new regime from its full impact:

• Only 50 per cent of foreign income arising in 2025-26 will be liable, whether held personally or in trusts.
• Offshore capital gains will be rebased to 2019 prices, meaning only gains realised since 2019 are liable. No time limit has been announced.
• A two-year window from 2025-26 to 2026-27 will allow UK residents that are ineligible for the new regime to bring onshore previously realised (or ‘stockpiled’) FIG outside of trusts to the UK at a discounted 12 per cent tax rate. This allows individuals to transfer some or all of their offshore wealth to the UK at a lower tax rate during this period, rather than facing the complexity of remitting smaller amounts of FIG to the UK on a more regular basis at the marginal tax rate.

3.14 The inclusion of the three ‘protections’ gives the costing an uneven profile over the forecast period, raising an average of £3.1 billion a year between 2026-27 and 2028-29. It reaches a steady state in 2030-31, when the yield is around £3 billion a year.

Table 3.3: Costing of the reform to the current non-domicile regime

3.15 The tax base consists of the estimated 5,500 individuals ineligible for the new regime who would pay full UK tax on their FIG, whether it is remitted to the UK or not. This is offset by the transitional protection to foreign income in 2025-26 and the rebasing of foreign gains to 2019. There are smaller losses from the tax currently paid by RB users on remitted FIG, from RB charges, and from the broadening of OWR. Estimating the tax base is highly uncertain: many non-domiciles on the RB opt in and out on a year-by-year basis, making it difficult to project future trends. The tax base relies on international agreements on the automatic exchange of information, which adds an additional layer of uncertainty. From 2026-27 onwards, we estimate the static costing generates £5.3 billion a year in additional income tax, NICs and CGT.

3.16 The behavioural response to the measure reduces the yield by an average of 40 per cent. This is the net impact of several very uncertain effects: the extent that those ineligible for the new regime use tax-planning to offset the increase in liability or whether they choose to emigrate; the amount of foreign income and gains that are forestalled due to the protections; and the take-up of the window to onshore stockpiled FIG, which raises CGT and income tax receipts in 2026-27 and 2027-28 but lowers them thereafter.

3.17 Three of the most uncertain and material behavioural responses are around:

• Tax planning: Outturn data following the previous 2017 reforms to this regime suggests that the average tax that was paid did not substantially change as a result. This is a more restrictive measure, but that is offset by the generosity of the protections, which mitigates the incentive to tax-plan. We have scaled down tax receipts outside the FIG window by 30 per cent to account for tax planning.
• Migration: Non-domiciled and/or high-net worth individuals are already highly transient, with 90 per cent having left the UK after seven years (when remittance charges take effect). But qualitative research finds that tax is not generally a first-order consideration in location decisions for high-net worth individuals, and that individuals are relatively ‘sticky’ due to personal and professional connection to an area.^[10] This is backed up by the evidence from the 2017 reforms, suggesting the numbers that exited post-reform were very low. This is another area where the impact is probably limited by generous protections. We assume that between 10 per cent and 20 per cent of current non-domiciles that are ineligible for the new regime leave the UK.^[11]
• Take-up of the FIG window: There is no reliable estimate of the amount of FIG that will be moved into the two-year window and this is a source of considerable additional uncertainty.

High-income child benefit charge

3.18 The high-income child benefit charge (HICBC) is a tax charge that applies to households where the income of the highest earner (the ‘lead’ income) exceeds a specified threshold, which has been fixed at £50,000 since the HICBC came into effect in 2013. Beyond the threshold, the household receives the same amount of child benefit but that is then tapered away by the charge, with the value eventually reaching 100 per cent of their child benefit award at an income of £60,000.

3.19 In this Budget the Chancellor has announced that, from April 2024, the HICBC income taper range will now rise to £60,000 to £80,000. This means that every household where the lead income is between £50,000 and £60,000 will no longer be subject to the HICBC and can effectively retain all their child benefit award. Those households with a lead income between £60,000 and £80,000 also stand to gain, though by gradually reducing amounts, as they are now subject to a less-than-100 per cent taper rate.

3.20 Table 3.4 sets out the three components of the costing. The static costing is £0.3 billion a year, due entirely to lower HICBC liability from households with lead incomes between £50,000 and £80,000. The direct behavioural effect of the policy is to increase take-up of child benefit. Overall, this measure is expected to boost the numbers registering for child benefit and paying the charge across the £50,000 to £80,000 income range. This increases child benefit expenditure by £0.5 billion in 2028-29, which is partially offset by £0.2 billion from higher receipts. We expect that most of the increase in take-up will occur in the £50,000 to £60,000 lead income range, as they no longer need to engage with the self-assessed income tax regime in order to claim child benefit payments.

3.21 A second set of indirect behavioural responses relates to labour supply.^[12] The changes to HICBC affect individuals’ work incentives through the impact on marginal tax rates and net weekly incomes, as set out in Box 3.2.^[13] The net impact is to increase average hours worked by existing employees by an amount equivalent to 10,000 FTEs. Overall, dynamic labour supply effects increase tax receipts by £0.2 billion a year, reducing the overall cost of the measure by a third.

Table 3.4: Costing of the high-income child benefit charge

A new duty on the e-liquids used for vaping

3.22 The Government has announced a new tax on the e-liquids that are used for vaping (effective from 1 October 2026) and a ban on the sale and supply of disposable vapes “as soon as possible”. The new tax applies a specified fixed amount of duty to a defined quantity of e-liquid and will be charged to UK manufacturers or importers. There are three rates that are differentiated according to the amount of nicotine content in the e-liquid.

3.23 It means that a vaping product that falls within the highest nicotine category and currently retailing at £3.00 (for 10 millilitres of e-liquid) will be subject to a £3.00 rate of duty (in 2026-27), raising its post-tax price to £6.60 (since we assume all of the duty will be passed through to consumers). We estimate vaping duty will raise £0.5 billion by 2028-29, though there are several uncertainties in the costing.

3.24 The tax base for this measure is the aggregate volume of consumed e-liquids, split according to the three different duty rate categories. Its estimate is one of the sources of uncertainty in this costing. There is incomplete data on the current number of UK manufacturers and importers, therefore the tax base is estimated using a combination of survey data, HMRC intelligence and external data sources. Further, the extent of existing non-compliance is also uncertain but is estimated usingcomparable existing tax gap estimates for tobacco products.^[14] Finally, the impact of the ban on disposable vapes will reduce the tax base to an uncertain extent. We assume that 40 per cent of existing users will switch to alternative vaping products and therefore be liable for vaping duty.

3.25 The second major source of uncertainty in this costing centres around several possible behavioural responses. These include:

• Reformulation: some producers might respond to the new duty by lowering the nicotine content in their products to drop to a lower duty rate, especially those whose current products are just above the duty thresholds.
• Forestalling: some businesses will boost their existing inventories to avoid the duty (but still benefit from the higher post-tax price). The extent of the forestalling will be limited by the products’ shelf-life (though e-liquid is relatively durable) and available storage capacity.
• Consumer demand: we assume that 100 per cent of the new duty will be passed-through to higher retail prices, which will lead some to lower their consumption, others to ‘downtrade’ to cheaper products or switch to tobacco, while some will stop vaping entirely. These effects are estimated using a combination of own- and cross-price elasticities and survey evidence. The announced rise in tobacco duties will mitigate some of the switching to tobacco products.
• Additional non-compliance: as discussed above, this is a sector with relatively high levels of existing non-compliance within its current regulatory regime, so it would not be surprising if that non-compliance was transferred to the tax system, and indeed it might increase in response to the new duty. To help mitigate against this, HMRC is receiving additional funding to support compliance in this area.

3.26 Overall, we have estimated that the combined impact of these behavioural responses is to lower the static yield by around 30 per cent.

The carbon border adjustment mechanism

3.27 On 18 December 2023 the Government announced plans to implement a carbon border adjustment mechanism (CBAM), effective from 1 January 2027. The CBAM is a new tax on certain carbon-intensive goods that are imported into the UK. CBAM is payable when the carbon price in the country of origin (if any) is lower than the carbon price paid by UK producers. It is designed to prevent the leakage of carbon emissions across international borders and follows the EU’s CBAM scheme which will be implemented from January 2026. The precise policy details include:

• The CBAM is a tax on the part of carbon (and carbon equivalent) emissions that are embodied in imports from the following sectors: aluminium, hydrogen, cement, fertiliser, ceramics, glass, and iron and steel.
• The liability is determined by the difference between the effective carbon price in the country of origin and that in the UK.^[15] It is calculated on a £ per tonne of CO2 equivalent basis.
• CBAM-liability will reside with the importer, above a yearly £10,000 de minimis.

3.28 Our central estimate is that CBAM will raise £0.2 billion by 2028-29, with the vast majority of that coming from just two sectors – iron and steel (predominantly from China) and aluminium. The costing assumes that EU carbon prices will continue to remain above those in the UK, meaning there is currently zero yield from EU imports, which are the largest single source of affected goods. This is one of several sources of uncertainty in this costing as the UK carbon price has proven to be highly volatile. If UK carbon prices were around £16 higher than the EU’s (last seen in October 2022), we estimate an extra £0.4 billion would be raised from charging EU imports in the UK.

3.29 There is also uncertainty over the size of the tax base which is the carbon content that is present in UK imports within the affected sectors. The carbon intensity of goods has been trending downwards, in line with countries’ net zero ambitions, but the precise future path of carbon intensity within CBAM-liable sectors is uncertain.

3.30 A third source of uncertainty in this costing is around the scale of the behavioural response, particularly since there is no observed evidence from an existing CBAM regime to base it on. There are several potential behaviours (which reduce the costing by around a third):

• The policy is designed to incentivise a positive behavioural change by producers and determining how successful that might be is difficult to predict.
• There is also uncertainty around the level of compliance, both in terms of how successful and timely the implementation of CBAM will be as well as the scope and incentive for those affected to attempt to circumvent the tax.
• It is also not known how other countries might respond to the CBAM. For example, some might introduce or increase their own carbon taxes, either to lower domestic emissions and/or to effectively reclaim taxing rights.

HMRC debt collection

3.31 The Government has announced a further set of HMRC compliance and debt collection measures in this Budget. The highest yielding measure places HMRC tax debt with private sector debt collection agencies using £140 million of additional funding, raising a cumulative £4.3 billion over the forecast period.^[16] It targets the recent growth in tax debt, with Chart 3.1 showing that both the flow of tax debt (the bars) was relatively stable until the pandemic. It then rose sharply in 2020-21 and has remained high since. The debt available for collection (the line, showing the starting stock plus the flow) reached £156 billion in 2022-23, which is almost two-and-a-half times larger than the pre-Covid average. The inflow of new tax debt has also remained high in the years following the pandemic, with 2022-23 being the highest-to-date at £117 billion.

Chart 3.1: HMRC tax debt available for collection

Measures with highly uncertain costings

3.32 We assign an uncertainty rating to all certified policy costings.^[17] The measures that we have given a ‘high’ or ‘very high’ uncertainty ratings are set out in Table 3.5. Most of the measures are revenue-raisers, whereas the cost of the two largest tax cuts in this Budget – the cut in NICs and freezing fuel duty – have a relatively low level of uncertainty.

Table 3.5: Costings of measures with high degrees of uncertainty

Behavioural responses to tax policy

3.33 Behaviour is a key component in tax costings and captures the ways individuals and businesses change their actions in response to the policy, thereby changing the tax base to which the new regime will be applied. Chart 3.2 displays the scale of the estimated behavioural response for selected tax policy announcements from this Budget. It shows the percentage of the static costing that remains once behavioural effects are factored in. It shows that:

• Some measures are expected to result in proportionately little behavioural response. For example, the demand for fuel is relatively inelastic with respect to its price, resulting in only a marginal reduction in the static cost of freezing fuel duty rates via increases in fuel purchases. Similarly, the cut in NICs has a relatively low direct behavioural response (fewer tax-motivated incorporations) though that is augmented by a larger indirect behavioural response (increase in the labour supply).
• The three excise duties – alcohol, tobacco and the new vaping duty – have different characteristics that lead to a variation in the respective size of their behavioural effects. The impact of tax changes on demand is measured by price elasticities and each sector has a degree of product substitutability within it (measured by cross-price elasticities). But alcohol is harder to forestall than the others and has lower levels of non-compliance. Relatively high duty rates on tobacco, especially cigarettes, suggests it may be at or beyond the point of revenue maximisation.
• The reform of the current non-domicile regime and the new carbon border adjustment mechanism are both discussed in detail above. Their costings are reduced by about a half and a third respectively on account of the behavioural response, with several very uncertain aspects in each costing.
• The cut in the higher rate of capital gains tax on residential property disposals from 28 to 24 per cent cumulatively raises tax revenues over the forecast period, through two channels. First, we expect that lowering the rate will ‘unlock,’ and so bring forward, some qualifying property disposals. Second, we expect the lower rate to have a small but permanent positive impact on the level of property transactions. Most of the additional tax from this measure relates to higher SDLT receipts. The unwinding of the first behavioural effect means that, by 2028-29 (the year shown in Chart 3.2), the overall effect is a small net cost.

Chart 3.2: Scale of behavioural response for selected tax policy changes

3.34 Those highly uncertain measures that are not discussed above are:

• From January 2026 the Government will implement the crypto-asset reporting framework (CARF), an OECD initiative which will require some crypto-asset service providers to report to HMRC on UK tax residents’ crypto-asset activity outside of the UK. Several aspects of the costing are uncertain, including the size of the tax base, the growth of crypto-assets and the extent that this measure improves compliance. The deterrent effect of CARF is particularly uncertain.
• The Government has extended the energy profits levy for a further year into 2028-29. There are two sources of uncertainty with this costing: the path of future energy prices remains uncertain while the impact of the measure on investment remains somewhat ambiguous. The investment allowance acts as an incentive while the continued instability of the tax regime acts as a disincentive.
• The Budget also contains measures that extend the generosity of the creative tax reliefs, whose cost has regularly exceeded expectations. The main uncertainties relate to the scale of the behavioural effect, especially the additional qualifying investment that is induced. There is also uncertainty around the level of non-compliance.

Policy measures not on the Treasury scorecard

3.35 Our forecasts include the effect of several policy decisions or related impacts that the Treasury has chosen not to present on its scorecard. The net effect of these is shown in the ‘non-scorecard’ line in Table 3.1 and can be explained almost entirely by two components:^[18]

• The Supplementary Estimates for 2023-24 that were laid in February 2024 and whose net effect is to reduce expenditure by £0.3 billion that year.
• An update to the levels of assumed spending beyond the Spending Review period that lower resource spending by an average of £0.8 billion a year.

3.36 The remaining non-scorecard decisions largely relate to fiscally neutral measures, such as the additional funding for the NHS in 2024-25 and local authority measures including council tax referendum limit waivers that have been granted to four councils, a change to council tax police precept and additional funds within the local Government settlement.

Update on previous measures

3.37 We cannot review and re-cost all previous measures at each fiscal event, but we do look at those where the original (or revised) costings seem to be under- or over-performing, and costings that were identified as particularly uncertain.

Personal tax policies and their effects on fiscal drag

3.38 The past three years have witnessed a sequence of large changes to the rates and thresholds of income tax and NICs (see Box 3.2 for the policy details and their impacts on labour supply). Freezing tax allowances and thresholds, rather than raising them in line with inflation, increases tax receipts as rising nominal wages tip more workers into paying tax, or paying a higher rate of tax, than would otherwise be the case. This ‘fiscal drag’ effect is especially pronounced when inflation is high, as it has been over the past 3 years. Table 3.6 sets out the actual and counterfactual thresholds and the associated rate of CPI inflation.

Table 3.6: Counterfactual and actual personal tax thresholds

3.39 Table 3.7 sets out estimates of the number of individuals who face a higher marginal rate of income tax as a result of fiscal drag taking them into a higher tax threshold over the next five years. By 2028-29, there are expected to be around 3.7 million more taxpayers overall, 2.7 million more higher-rate taxpayers, and 600,000 more additional-rate taxpayers than if all allowances and thresholds had been indexed to inflation, and the additional rate kept at £150,000.

Table 3.7: Effect of threshold freezes on additional taxpayers

3.40 The net fiscal impact of the personal tax threshold freezes and NICs rate cuts announced since March 2021 has been to increase tax receipts by £19.7 billion by 2028-29. As shown in Table 3.8, this comprises:

• Changes to thresholds that raise receipts by £41.1 billion by 2028-29. This is primarily driven by £33.6 billion of revenue from freezing the income tax personal allowance and higher rate threshold since March 2021, relative to raising them by CPI. This is £8.2 billion more than we forecast in our March 2023 EFO, as higher and more persistent inflation has increased the difference between the frozen thresholds and the rates to which they would have risen absent the measures.
• A £21.4 billion cost from the successive cuts in the rate of NICs announced in November and in this Budget, each costing £10.7 billion in 2028-29.

3.41 Overall, the personal tax cuts announced in the 2023 Autumn Statement and 2024 Spring budget reverse around a half of the total additional tax revenue raised from the freezing of thresholds in 2028-29.

Table 3.8: Latest costings of personal tax measures

Box 3.2: The labour supply impacts of personal tax policies

Between March 2021 and November 2022, almost all the main allowances and thresholds in income tax and national insurance contributions (NICs) were frozen rather than indexed to inflation, as is the default in nearly all cases, up to and including 2027-28. The Chancellor has subsequently announced personal tax cuts that are due to offset around a half of the resulting impact on the personal tax burden by 2028-29. Since March 2021, the main policy changes have been:

• Freezing key income tax thresholds: the personal allowance (PA) and higher rate threshold (HRT) at £12,570 and £50,270 from 2022-23 to 2027-28.
• Lowering the additional rate threshold (ART) from £150,000 to £125,140, from April 2023, to align it with the PA taper.
• Various other changes to thresholds and allowances: the increase in the NICs primary threshold and lower profits limit to align with the PA and then freezing both to 2027-28. The NICs upper earnings limit, upper profits limit and employer secondary threshold were also frozen to 2027-28.
• November 2023 NICs cuts: a 2 percentage point cut in the main rate of employee NICs (effective from 6 January), a 1 percentage point cut in the main rate of NICs for the self-employed, and the removal of the requirement to pay Class 2 NICs (both effective from 6 April) announced in the Autumn Statement.
• March 2024 NICs cuts: a further 2 percentage point cut in the main rate of both employee and self-employed NICs (from 6 April) announced in this Budget.

Table A shows our estimates for total effects of these policy decisions on work incentives, given high inflation, over the period from 2021-22 to 2028-29.^a Some of these effects have already occurred while others will continue over the forecast period. The policies have both positive and negative effects on work incentives. But their net effect is to increase labour supply by an estimated 79,000 in full-time equivalent (FTE) terms (0.2 per cent) by 2028-29. Almost all of the increase comes from existing workers choosing to work longer hours (78,000 FTE), while the net impact from additional employment is 1,000 FTE. Including the impact of the government’s recent childcare expansion and welfare reforms, we estimate the overall net increase in total labour supply by 2028-29 is 193,000 people in full-time equivalent terms.

Table A: Labour supply impacts of major personal tax policies

Looking at the labour supply impacts of each of these policies in turn:

• Fiscal drag from frozen thresholds given high inflation will directly affect the work incentives of the 39 million taxpayers that will earn above the personal allowance by 2028-29. But the effect is driven by the 7 million subset of workers that will be newly taken into tax or into higher tax brackets, and so reduce their hours in response to facing significantly higher marginal tax rates than would otherwise be the case. This decreases labour supply by an estimated 130,000 in FTE terms by 2028-29, relative to a counterfactual in which thresholds rose in line with inflation from 2022-23 to the values shown in Table 3.6.^b We judge that 55 per cent of this impact has already occurred, so the effect over our forecast period is less than half the total, whereas the positive NICs cuts’ effects are all to come.
• The cuts to NICs announced in this Budget change the work incentives of the 28 million employees and 2 million self-employed individuals that currently earn between the primary threshold and upper earnings / profits limit. As described above, this operates in a similar way to November’s announcement, generating a labour supply impact of 98,000 in FTE terms, with a combined impact across both measures of 199,000 FTEs.^c
• The HICBC reform has very large effects on marginal post-tax earnings, but for relatively small numbers of people. The main beneficiaries are the 170,000 individuals with a lead household income between £50,000 and £60,000. As described above, this measure boosts labour supply by 10,000 in FTE terms.

Focusing just on the most recent tax changes (the November NICs cuts, the NICs cuts in this Budget and the changes to the HICBC), whose effects will be felt over the next five years, plus the remaining impact of the frozen tax thresholds, the net impact from now to 2028-29 is a positive 150,000 FTE workers.

These estimates are subject to considerable uncertainty, as are those for the welfare, childcare, and other measures announced in the last two fiscal events (as discussed in Box 2.1) that increase labour supply by 114,000 in FTE terms. These sources of uncertainty include:

• The responsiveness of individuals in different demographic subgroups and income bands to changes in financial incentives. Plausible alternate assumptions for the elasticities used in our modelling could substantially impact our estimates.
• Another source of uncertainty relates to our underlying economy forecast. This forecast reflects the increase in the size of the workforce that is implied by the latest labour force and population projection releases. As a result we have increased our previous estimate of the impact of the November 2023 NICs cuts from 94,000 to 101,000. Deteriorating economic conditions could also reduce our estimates, for instance, if spare capacity opens up over the coming years as labour demand fails to match supply.

a) These are derived by inputting our own assumptions about elasticities and other parameters into the Treasury’s labour supply model to assess the impacts of personal tax policy changes on work incentives. For further details on the methodology, see OBR, The labour supply effects of the Autumn 2023 National Insurance Contributions cut, February 2024.

b) We use lagged, third-quarter CPI as a close proxy for the previous year’s September CPI figure with which most thresholds are uprated. This therefore essentially reflects the impact of not raising all the thresholds shown in Table A with inflation, as well as leaving frozen some thresholds that have been fixed for a longer period, like the Personal Allowance Taper. For the Additional Rate Threshold (ART) of Income Tax, which was lowered from £150,000 to £125,140 in 2023-24, we started assuming divergence from the counterfactual from that point, to focus on the impact of decisions to freeze tax thresholds (or, in some cases, implicit decisions to leave them frozen), and avoid conflation with other contemporaneous policy changes. We have adjusted changes in non-devolved taxes to the UK level by simple scaling, abstracting from some small differences in uprating policy in Scotland.

c) As described below, we have made an upward adjustment to the labour supply effect estimate of the November 2023 NICs cut.

Updates on other measures

3.42 We have also reviewed the following measures:

• The VAT Retail Export Scheme enabled non-EU visitors to the UK to reclaim the VAT incurred in purchasing goods from participating retailers. It was abolished following the UK’s exit from the EU in January 2021. Box 3.3 reviews our original costing.
• Tariffs on electric vehicles (EVs). The Government has announced a three-year extension to the existing trade rules governing the trade in EVs between the UK and EU, delaying the planned introduction in the rules of origin requirements until 1 January 2027, reducing customs duties by £0.2 billion a year. We previously assumed that traders would be ready for the changes by 2024 and that trade in EVs with the EU would then be effectively tariff free. We make an equivalent increase in our pre-measures customs duties forecast to reflect those receipts that would have been collected by traders not meeting the rules of origin requirements.
• The ‘Additional Information Form’ (AIF) requirement for claiming R&D tax relief, also known as ‘G-forms’, became mandatory from 8 August 2023. This requires claimants to provide additional details on the company, the agent and the R&D activity. We estimate that AIFs will initially reduce small and medium enterprises’ (SMEs) R&D tax credit claims by 33 per cent and by 10 per cent a year over the medium term, as agents become more familiar with the new requirements. This reduces R&D tax credit expenditure by £0.7 billion in 2023-24 and by an average of £0.5 billion a year subsequently. These forms have only been in operation for a matter of months and there was some forestalling of claims ahead of its introduction, so these estimates are uncertain. There is no evidence of large business’ R&D relief claims being affected.

Box 3.3: VAT Retail Export Scheme: review of the 2020 policy costing

In November 2020, the Government announced that the VAT Retail Export Scheme (RES) was to be closed from 1 January 2021, following the UK’s departure from the EU. The scheme allowed non-EU visitors to the UK to claim back the VAT incurred on the purchase of certain goods in the UK from participating retailers, subject to conditions. These included that the goods must be purchased from a participating UK retailer, the purchaser needed to complete and submit a paper form, and the goods needed to be exported from the EU within three months of purchase. In practice, the vast majority of claims were processed at departing airports.^a

The scheme’s refunds largely related to spending in a relatively small number of luxury stores, such as those located in central London and Bicester Village. Closing the scheme aligned the VAT treatment of non-EU visitors with those from the EU, complying with WTO rules on non-discrimination. In this review we look back at the original November 2020 costing with the benefit of the latest available data and report our revised estimate of the fiscal impact in 2025-26, which was the final year of the scorecard costing. Table B provides a side-by-side comparison of the main costing elements from the 2020 estimate and our current view, both relating to 2025-26.

Table B: Comparing the original and revised estimates of the costing

The costing can be broken down into three main elements:^b

• The static costing is simply the Exchequer saving from no longer incurring the cost of the VAT refunds.
• Some non-EU visitors will be deterred from visiting the UK following the abolition of the scheme. This direct behavioural effect was captured in the original costing, with the post-measure VAT RES spending (and associated tax receipts) from this group dropping to zero. The original costing did not include any indirect behavioural effects, namely that there would be a further loss in receipts from the non-VAT RES spending that this group would otherwise have undertaken.
• For those VAT RES shoppers that do still come to the UK, the effective rise in price leads them to spend less on these goods, lowering tax receipts. This direct behavioural response from this group is included in the original costing. But it did not factor in the indirect behavioural effect, which in this case refers to previous VAT RES spending by this group that is diverted to non-VAT RES items, which partially offsets the impact of lower VAT RES spending.
• A further potential channel for indirect effects is the possibility that lower external demand for VAT RES goods might reduce employment and long-run potential output.

Reviewing the static costing

The tax base for this costing is the pre-measure level of spending on VAT RES goods by non-EU visitors. The static costing is estimated using HMRC evidence on 2019 refunds, which suggests that just over £520 million of VAT was refunded that year to 1.2 million visitors, implying that spending on VAT RES goods was around £2,700 per person who claimed a refund.^c

This is projected over the forecast period using RPI-inflation and the expected number of non-EU visitors. The measure was introduced during the pandemic, when visitor numbers were historically low, so the relatively slow expected uptick in visitors meant the original static costing only returned to £535 million from 2023-24, rising to £617 million by 2025-26.

In the event, the pandemic suppressed visitor numbers for longer than we expected but numbers then rebounded more sharply so that, in cumulative terms, visitor numbers are now expected to be slightly higher (4 per cent) by 2025-26 than we thought in November 2020. By contrast, cumulative CPI inflation (which we decided to switch to for this review since we think it is a more appropriate measure) to 2025-26 is materially higher (16 per cent) than our November 2020 forecast.^d This higher inflation is the main reason that in the reviewed costing we raise our estimate of the static costing to £685 million.

Reviewing the direct behavioural impact from non-EU visitors no longer travelling to the UK

The 2020 costing estimated that the policy change would result in 30,000 fewer visitors travelling to the UK in 2025-26. Of the 1.2 million visitors that submitted VAT RES forms in 2019, it was assumed that just under 50 per cent were holiday travellers (with the majority of the remainder visiting for business or family reasons).^e For travellers from the rest of the world, the costing estimated an increase in the overall cost of their UK visit of around 3 per cent,^f then applied a price elasticity to derive the reduction in visitor numbers.

The elasticity was based on the one UK-specific tourism estimate known to us at the time, which estimated an elasticity of -1.28.^g To reflect that VAT RES travellers will see larger changes in visit costs and may be more price sensitive, we scaled up the elasticity by around 50 per cent to -1.9, though this was an uncertain judgement. The tax receipts that would have accrued from this group’s spending on VAT RES goods reduces the savings from the measure in the 2020 costing by £12 million in 2025-26.

In reviewing this direct behavioural assumption, we have considered trends in actual visitor numbers since the policy was introduced. We have also reviewed external estimates of the price sensitivity of travel and tourist expenditure.

Visitor numbers are driven by a wide range of factors, and the uncertainties during this period are amplified by both the pandemic and the UK’s exit from the EU. The total number of UK visitors has now returned to close to pre-pandemic levels, as has the share from both EU and non-EU visitors (Chart C (left panel)). There has been a moderately bigger bounce-back in the number of visitors from the United States to some other European countries (Chart C right panel). However, the overall difference is not large, and we have not seen a significant change in the composition of UK visitors (between country of origin or reason for travel).

Our review of external estimates of the price sensitivity of international travellers points to a range of -2.1 to -0.4, so the elasticity used in the 2020 costing is toward the top of the range.^h We have made a small adjustment to account for those visitors who, though stating they were primarily travelling for non-holiday reasons, might have also been motivated by VAT RES shopping – this takes that proportion from just under 50 per cent to just over. Our revised estimate therefore is that the policy leads to 34,000 fewer VAT RES claimants by 2025-26 and this lowers receipts by £15 million in 2025-26, £3 million more than in the 2020 costing. As groups of VAT RES travellers will see proportionally greater changes in visit costs than other travellers from the rest of the world, there is particularly significant uncertainty around this figure. So, while the data has not cast doubt on our original assumption, there is significant uncertainty around it, and it is possible that the number of deterred visitors could be higher or lower than we assumed. For the resulting behavioural cost to the Government to outweigh the policy’s static savings, this group’s responsiveness would need to be very significantly higher than is suggested by any of the evidence we have considered.

Chart C: Number of visits to the UK by overseas residents

Reviewing the indirect behavioural impact from non-EU visitors no longer travelling to the UK

There is also a loss of tax that would have been collected on spending by deterred visitors on non-VAT RES goods and services. We did not explicitly account for the fiscal impact of this indirect effect of the policy in our November 2020 forecast. However, as in every forecast, the effect on overall aggregate demand of the full policy package (including VAT RES abolition) would have been captured in our economy forecast (via the fiscal multiplier framework discussed in Box 2.1). We would not normally separately quantify the indirect demand effect of a measure such as this, which is relatively small in terms of the UK economy, but we have done so for this review to illustrate its scale.

To do so, we assume that, on average, each deterred visitor was travelling with one companion and that, based on the 2019 data, spending on non-VAT RES items that year was around £1,750 per person (though this estimate is very uncertain). Taking these together, and growing them in line with CPI-inflation and passenger numbers over the forecast period, we estimate that reduced spending by deterred visitors on non-VAT RES goods reduces tax receipts by a further £24 million in 2025-26.

Reviewing the behavioural impact from reduced spending by non-EU visitors that still travel to the UK

The policy raises the net effective price of VAT RES goods for those non-EU visitors that still travel to the UK. The 2020 costing expected this group to respond by reducing the amount they spend on these items. The methodology used to estimate this assumed that VAT RES spending by this group would fall by 29 per cent (around a £785 per-person drop in 2019 terms), closer to (but still above) EU-visitor spending levels. This resulted in a £139 million loss in receipts in 2025-26. Chart D shows that the ratio of non-EU to EU spending has indeed fallen since 2021, though the extent to which this is a direct consequence of the policy, rather than due to other factors, is not determinable from the data.

A price elasticity of demand approach is a more appropriate way to estimate this behavioural response. The VAT RES demand response from this (undeterred) group should more accurately reflect the change in the price of tourism goods, rather than the overall cost of their trip. Based on the available literature,ⁱ the -1.8 intensive margin elasticity implied by a 29 per cent reduction in spending is arguably high. But the literature does not generally consider luxury shoppers explicitly who we would expect to be relatively more price sensitive. And the price rise due to the policy will lead some shoppers to switch from VAT RES to non-VAT RES spending (this effect was not explicitly accounted for in the 2020 costing).

There is little empirical evidence to determine what proportion of this spending will be displaced. But as many luxury goods are not close substitutes for other goods and services consumed by holidaymakers, we have made the judgement that only a third of the 29 per cent reduction in VAT RES spending will be diverted elsewhere. This lowers the implied elasticity to -1.2 which we believe is more central. Overall, we have revised down the total loss in receipts from those non-EU visitors that still travel to the UK to £102 million in 2025-26, £36 million lower than the 2020 costing, largely due to including the impact of spending that is displaced to non-VAT RES sectors.

Chart D: Per-visit spending levels of all UK visitors

Potential wider demand and supply-side indirect behavioural effects

This measure has clearly had significant impacts on the gross sales of affected businesses. As mentioned above, to assess the potential wider demand impact of policy measures on our economy forecast in November 2020 we used our ‘fiscal multipliers’. These assume that any near-term changes in demand from tax increases (like this one) taper away to zero as exchange rates, monetary policy, and real wages adjust to bring output back in line with its potential level and inflation back to target. To fully account for the transmission of changes to the gross sales of affected businesses to the wider economy we would also need adjust for the high import content of many luxury goods, which would significantly reduce the impact on net external demand.

We would only assume changes in fiscal policy have permanent impacts on the level of real potential GDP, the supply side of the economy, when strong evidence suggests that the effects on the value added in the UK are significant, durable, and additional.^j With unemployment having remained broadly in line with our estimate of its neutral rate and with vacancies high, this measure is more likely to have reallocated employment and activity between the UK’s sectors and regions.

Conclusion

On review, our 2020 methodology still appears reasonable, and our costing to have been a central estimate (although one surrounded by considerable uncertainty). We have updated it to account for the impacts of visitors’ spending on non-VAT RES spending, but continue to believe that this measure is unlikely to affect significantly the productive capacity of the economy.

We have focused on VAT RES rather than airside shopping because of its greater fiscal costs. Based on our own assessment and from responses received from external stakeholders the behavioural uncertainties around VAT RES are more significant for our forecast than those surrounding airside shopping. But both measures operate via similar channels, and so our main analytical conclusions are also likely to also be relevant for airside shopping.

Table C: Revised costing of the abolition of VAT RES

a) Around 95 per cent of claims were submitted at major London airports.

b) For more detail on policy costings, see our Briefing paper number 6: Policy costings and our forecast, March 2014.

c) A single traveller might be claiming for one or more other members of their travelling party.

d) Switching from RPI to CPI inflation is not material to the estimate: RPI growth was 21 per cent compared to CPI’s 16 per cent.

e) Based on responses to the International Passenger Survey.

f) Based on assuming the cost is 100 per cent passed through to higher VAT RES prices, many of the commercial companies that processed VAT refunds charged an administrative fee for their services. The costing accounts for this by using the net effective rise in prices. Not all participating retailers applied an administrative fee.

g) PWC, The Impact of Taxes on the Competitiveness of European Tourism – Final Report, October 2017

h) See, for example, Scottish Government, Elasticities relevant to tourism in Scotland: evidence review, June 2021; and Home Office, A review of evidence relating to the elasticity of demand for visas in the UK, March 2020.

i) See, for example, Scottish Government, Elasticities relevant to tourism in Scotland: evidence review, June 2021

j) For more detail, see our Briefing paper No.8: Forecasting potential output – the supply side of the economy, November 2022.

Policy risks

3.43 Parliament requires that our forecasts only reflect current government policy. As such, when the Government sets out ‘policy ambitions’ or ‘intentions’, we ask the Treasury to confirm whether they represent firm policy. We use that information to determine what should be reflected in our forecast. Where they are not yet firm policy, we note them as a source of risk to our central forecast. A full database of risks to this forecast and changes from previous updates is available on our website. Here we summarise risks that have changed materially since our March forecast and those that are new.

3.44 The risks that have crystallised since our previous forecast include two new taxes, on vaping duty and CBAM, that are described above. A third is investment zones (IZs), with the Government announcing in this Budget several tax reliefs and business rate retentions that will be offered to the eight IZs across England whose effects are included in our forecast. We have yet to include the fiscal impact of non-English IZs.

3.45 Policy risks that have evolved or that are new since November include:

• The Government has announced additional compensation payments to the victims of the Post Office Horizon scandal. Government figures show that around £160 million has been paid to date to over 2,700 claimants across the three compensation schemes. The Government has also introduced legislation exonerating those who were wrongly convicted based on the Horizon evidence, which will entitle them to at least £600,000 in additional compensation. The final cost of the compensation pay-out will depend on the circumstances set out in individual claims, and we will include the impact in our forecast once this has been determined.
• Infected blood compensation payments. In July 2017, the Government announced an inquiry into the use of infected blood products used for transfusions in the 1970s and 1980s. In August 2022, the Government accepted the recommendations of the inquiry’s interim report to provide interim compensation payments of £100,000 to those currently in receipt of support from one of the four support schemes. The final report is due to published in May 2024, and we will incorporate the impacts of any further recommendations into our forecast once they are sufficiently certain.
• The Government has announced its intention to deliver a retail sale of part of its NatWest shareholding, subject to supportive market conditions and achieving value for money. There is insufficient detail at this stage, for example on the timing of the sale, to incorporate this into our forecast.
• The reform of the current non-domicile regime did not specify the treatment of inheritance tax, but the Government has signalled its intention to move to a residence-based system subject to consultation. We will include its effects in due course.
• The Chancellor has indicated plans to introduce a UK ISA, which will allow savers to invest up to £5,000 a year into UK companies (in addition to the existing £20,000 ISA limit). The design is being consulted on.
• In March 2023, the UK joined the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), becoming the first new member since the trade bloc was established in 2018 and the first European member. However, there is not yet enough policy detail for us to estimate the impacts that this deal may have on our forecast at this time. The Government has estimated the deal is expected “increase long-run UK GDP by around £2.0 billion each year”,^[19] which supersedes the estimate in our November EFO.
• Devolved administration policies. On 8 February 2024, the Deputy First Minister announced that the minimum unit price for alcohol in Scotland will rise from 50p to 65p when current regulations expire at the end of April. The Scottish and Welsh Governments have also decided to follow the UK Government in banning disposable vapes. We will incorporate the fiscal impacts of these decisions in our next forecast.
• The Government has announced a competition to select the best small modular reactor (SMR) technology and provide funding if the technology proved viable. As part of this, the Government is considering purchasing sites which could be used for SMRs or nuclear plants. The competition is ongoing and no final decision has been made. The Government has also announced a roadmap to increase nuclear power generation by 2050 which includes up to £300 million of investment. We will incorporate the impacts of these initiatives into our forecast once they are sufficiently certain and the timeline has been confirmed.
• This Budget continues the pattern of Chancellors’ not implementing their stated policies to raise fuel and alcohol duties. We estimate that the cumulative cost of freezing fuel duty rates between 2010-11and 2024-25, relative to increasing them in line with RPI inflation has risen to around £90 billion, after factoring in the expected negative impact on demand for fuel from higher duty rates. This includes the impact of the ‘temporary’ 5p cut to rates introduced in Spring Budget 2022, but which has now been extended to a third year at this Budget.

Chapter 4: Fiscal outlook

Introduction

4.1 This chapter:

• notes classification issues affecting our forecast (from paragraph 4.4);
• describes the outlook for public sector receipts (from paragraph 4.5) and public sector expenditure (from paragraph 4.44);
• presents forecasts for borrowing and other deficit aggregates, including measures of the overall, current, and primary fiscal balances (from paragraph 4.67);
• describes the outlook for financial transactions and balance sheet aggregates and for government lending to the private sector (from paragraph 4.72); and
• summarises key uncertainties and risks to the fiscal outlook (paragraph 4.81).

4.2 The forecasts in this chapter start from the estimates of 2022-23 outturn data published by the Office for National Statistics (ONS) in their 23 January public sector finance release. We then present an in-year estimate for 2023-24 that makes use of ONS outturn data for April 2023 to December 2023 and HMRC administrative data for tax receipts in January 2024. Finally, we present forecasts for 2024-25 to 2028-29. Throughout this chapter we compare our latest forecast with those from our most recent Economic and fiscal outlook (EFO) published in November 2023.

4.3 In common with recent EFOs, there is considerable uncertainty around this forecast in part reflecting the volatile external environment, notably from the Russian invasion of Ukraine and conflict in the Middle East. Important drivers of the fiscal forecast, such as the paths for inflation and interest rates, remain particularly uncertain, as do the impacts of migration, inactivity, and productivity on growth and the labour force. Throughout this chapter we discuss how changes in these factors have affected our forecasts. Chapter 5 also assesses the fiscal implications of some of these key sources of uncertainty.

Classification and other statistical changes

4.4 No ONS classification decisions since our November forecast have affected this forecast. At this event the Government has introduced a new tax on vaping and the carbon border adjustment mechanism. On the advice of Treasury classification experts and pending an ONS decision we are treating these as taxes on production.

Public sector receipts

Summary of the receipts forecast

4.5 In 2022-23, total public sector receipts as a share of GDP reached 40.3 per cent, a sharp 3.4 percentage point increase on the pre-pandemic level of 36.9 per cent of GDP in 2019-20. Receipts are forecast to then rise more slowly as a share of the economy, by a further 0.9 percentage points, to reach 41.2 per cent of GDP by 2028-29. The ratio of National Accounts taxes^[20] to GDP is forecast to decrease slightly this year. This is driven by the November 2023 cuts to the main rates of employee and self-employed National Insurance contributions (NICs), lower energy prices reducing oil and gas revenues, and the weak housing market reducing property taxes. Tax as a share of GDP then rises gradually in every year of the forecast to 37.1 per cent of GDP in 2028-29, which would be the highest level since 1948. This would be 4.0 percentage points above the pre-pandemic level of 33.1 per cent of GDP in 2019-20 (Chart 4.1).

4.6 Relative to our November forecast, tax as a share of GDP has been revised down by 0.2 percentage points in 2023-24 and, on average, by 0.5 percentage points thereafter. In the near term this is driven by weaker than expected receipts in recent months, in particular from corporation tax, VAT, and self-assessment income and capital gains taxes. In the medium term, the lower tax take is due, in roughly equal parts, to a less tax-rich economy forecast and to the policy measures announced in this Budget, in particular, cutting the main rates of employee and self-employed NICs by a further two percentage points and freezing fuel duty rates (and extending the 5p cut) for a year.

Chart 4.1: National Accounts taxes as a share of GDP

4.7 The tax-to-GDP ratio in 2028-29 is forecast to be 4.0 per cent of GDP higher than it was in 2019-20 (Chart 4.2), roughly three quarters of which (2.9 per cent of GDP) has already occurred between 2019-20 and 2023-24. The increase in the tax take over this period can be broken down into the following components:

• The pre-pandemic tax system would have increased the tax take by 0.7 per cent of GDP.
• Underlying forecast changes, excluding the direct impact of policy measures, increase the tax take by 1.9 per cent of GDP. Before the impact of policy, as a share of GDP: income tax receipts have increased because of a rise in the labour share of GDP and strength in earnings growth at the top of the income distribution; onshore corporation tax receipts have increased due to strong profits growth in higher-tax paying sectors; and VAT receipts have increased partly due to a reduction in the VAT gap.^[21]
• The direct impact of major tax increases raises the tax take by 2.0 per cent of GDP. This includes the threshold freezes for income tax and NICs, which increases the tax take by 1.3 per cent of GDP, and the increase in the corporation tax rate from 19 to 25 per cent, which increases the tax take by 0.6 per cent of GDP.
• The direct impact of major tax cuts reduces the tax take by 0.9 per cent of GDP. Personal tax cuts in the form of the reductions in (NICs) employee and self-employed main rates announced in November 2023 and in this Budget reduce the tax take by 0.7 per cent of GDP. Corporate tax cuts in the form of the super-deduction announced in March 2021, the temporary full-expensing announced in March 2023, and the permanent full-expensing announced in the November 2023 reduce the tax take by a further 0.2 per cent of GDP.
• The direct impact of all other tax measures announced since March 2020 increases the tax take by 0.4 per cent of GDP. This includes tax increases such as the increases in council tax, the Pillar 2 reforms to corporation tax and the introduction and extension of the energy profits levy and is partially offset by tax cuts such as subsequent freezes to fuel duty and increasing the primary NICs threshold.

Chart 4.2: The rise in the tax-to-GDP ratio between 2019-20 and 2028-29

Change in receipts since our November 2023 forecast

4.8 Relative to our November forecast, and including the impact of Budget measures, total public sector receipts are forecast to be £3.8 billion higher in 2023-24 and lower by an average of £13.1 billion a year between 2024-25 and 2028-29. Excluding fiscally neutral items that are offset in spending, receipts have been revised down by £1.2 billion this year and by an average of £17.2 billion a year between 2024-25 and 2028-29. Forecast changes, excluding fiscally neutral items, are driven in the short term by slightly weaker-than-expected overall receipts outturn this year, and in the medium term by a weaker forecast for inflation and the other nominal drivers of the key tax bases, in particular lower nominal earnings and consumption.

4.9 The direct impact of measures announced in this Budget reduces receipts by an average of £8.5 billion between 2024-25 and 2028-29. Tax-cutting measures raise borrowing by an average of £12.6 billion a year over the forecast, in particular the further reduction in NICs and cancellation of the planned fuel duty increase. The cost of these tax cuts is partially recouped in the medium-term by a set of revenue-raising tax measures, the most significant of which is the reform of the current non-domicile regime. The direct policy effects are partially offset by the indirect effect of the measures, largely from the additional revenue generated by the temporary boost to demand that results from the overall fiscal loosening due to the measures, and by the labour supply benefits of the NICs cut and the change to the high-income child benefit charge (HICBC).

Table 4.1: Public sector receipts: changes since November 2023

4.10 Risks around the receipts forecast come from both uncertainty about the macroeconomic outlook and the implementation of stated government policy. Chart 4.3 shows our forecasts of receipts as a share of GDP since 2010 compared to actual outturn (with forecast receipts based off predicted growth rates to account for revisions to GDP).

4.11 During the 2010s, receipts outturn was generally lower than we had forecast, which was largely due to our economy forecast overestimating productivity growth.^[22]
More recently, in our forecasts ahead of the pandemic, we generally under-forecast the medium-term receipts-to-GDP ratio. This is primarily due to the unexpected impact of the post-pandemic period of high inflation driving strength in nominal receipts, combined with policy decisions to raise tax rates and freeze personal tax thresholds. This illustrates some of the key risks facing our forecast at the current time, which are discussed further in this chapter and in Chapter 5, related to uncertainty around productivity growth, inflation, and future policy decisions.

Chart 4.3: Current receipts as a share of GDP, successive OBR forecasts and outturn

Box 4.1: The fiscal implications of tax gaps

The tax gap is the difference between tax collected and total theoretical tax liabilities (the amount of tax that should, in theory, be collected if individuals and businesses paid all tax due).^a In 2021-22, HMRC estimated the tax gap to be £35.8 billion, or 4.8 per cent of total theoretical tax liabilities (equivalent to 1.5 per cent of GDP), meaning that HMRC collected 95.2 per cent of all tax due. Changes in the tax gap represent a risk to our forecast: all else equal, a smaller tax gap will increase receipts, whereas a higher gap reduces receipts. This box explores how the tax gap has evolved since 2005-06, the main drivers of the changes, the assumptions we make about the tax gap in our forecast, and the associated uncertainties.

As shown in Chart A, as a share of GDP the tax gap fell from 2.3 per cent in 2005-06 to 1.5 per cent in 2021-22 (equivalent to a fall from 7.5 per cent to 4.8 per cent of tax liabilities). The main drivers of this reduction were:

• The VAT gap has reduced from 0.8 per cent of GDP in 2005-06 to 0.3 per cent in 2021-22 (equivalent to a fall from 14.0 per cent to 5.4 per cent of tax liabilities), explaining about two-thirds of the reduction in the overall tax gap. Much of this fall has occurred since 2013-14. HMRC have introduced several measures in recent years aimed at closing the VAT gap, and the reduction in the use of cash, especially since the pandemic, is also likely to have contributed.
• The tax gap related to personal income taxes has reduced from 0.7 per cent of GDP in 2005-06 to 0.5 per cent of GDP in 2021-22 (the equivalent of 4.5 per cent and 3.0 per cent of tax liabilities, respectively), which is likely to reflect HMRC activity to drive better compliance around business payroll. There has also been a reduction in the excise tax gap from 0.3 per cent of GDP in 2005-06 to 0.1 per cent in 2021-22 (the equivalent of a fall from 8.3 per cent to 6.1 per cent of tax liabilities).
• The corporation tax gap has increased from a low of 0.2 per cent of GDP in 2011-12 to 0.4 per cent in 2021-22 driven by the small business corporation tax gap, which increased from 8.8 per cent of tax liabilities in 2011-12 to 29.3 per cent of tax liabilities in 2021-22. The tax gap (across all taxes) attributable to small businesses has increased to 56 per cent of the overall tax gap in 2021-22. Aiming to address this, HMRC has introduced new small business support activity and in March 2023 announced a review of small business guidance and forms.

Chart A: Tax gap as a share of GDP

Comparisons with other countries are difficult because of limited availability of data and methodological differences, but where these are available the UK generally compares favourably. National Audit Office reporting of tax gaps from the US, Canada, Australia and Italy between 2011 and 2016 ranged from 7.4 to 19 per cent of tax liabilities, compared to between 5.5 and 7.0 per cent of tax liabilities in the UK during the same period.^b

In our baseline forecast we assume that the tax gap remains flat as a proportion of tax liabilities. This is consistent with the broadly flat overall trend seen since 2017-18. There are upside and downside risks to this assumption. Factors such as the ongoing digitalisation of the tax collection system and a further reduction in the use of cash in the economy could push down on the tax gap. On the other hand, subdued economic growth and cost of living pressures could lead to wider non-compliance. We do adjust our baseline forecast to account for specific measures directly targeted at reducing the tax gap. We scrutinise such measures to ensure the estimated yield is reasonable and accounts for behavioural responses, for example that some individuals or businesses will find alternative ways to avoid complying with new measures.

As an illustration of the risk to our forecast, if the tax gap, as a share of GDP, fell by a further 0.3 per cent, roughly half of the decrease in the four years between 2013-14 and 2017-18, that would increase receipts by £8.4 billion a year, on average.

a) The tax gap estimates only cover the taxes administered by HMRC, so they exclude any taxes and duties administered elsewhere, such as council tax, business rates and vehicle excise duty, as well as charges such as the congestion charge.

b) In Tackling the tax gap the NAO notes that these figures are not always directly comparable due to the different measurement methods and tax regimes included.

Tax-by-tax analysis

Income tax (excluding self-assessment) and NICs

4.12 Pay as you earn (PAYE) income tax and NICs receipts are expected to reach £421.0 billion in 2023-24, £30.6 billion higher than in 2022-23, driven by strong nominal earnings growth combined with frozen tax thresholds. This has been partially offset by the cut to NICs rates announced in November, which came into effect in January, which means NICs receipts are expected to be lower as share of GDP than in 2022-23.

4.13 From 2024-25 onwards, we expect receipts to grow, on average, by 3.1 per cent a year to reach £490.7 billion (15.3 per cent of GDP) by the end of the forecast period.^[23]
This primarily reflects our forecast for cumulative nominal earnings growth of 12.6 per cent over the rest of the forecast, coupled with the continued freezes to income tax thresholds that generate powerful fiscal drag (see Table 3.6). The effective tax rate (ETR) of income tax excluding self-assessment (SA) and NICs rises by 0.8 per cent by the end of the forecast period, slowing in the final year when thresholds are once again uprated with inflation.

4.14 Relative to our November forecast, non-SA income tax and NICs receipts have been revised down across the forecast. This is for the most part due to the cut to NICs announced in this Budget, which is partially offset by stronger outturn data in earlier years but gradually diminished by weaker nominal earnings growth. As shown in Table 4.2, the main sources of these changes are:

• On a pre-measures basis, by 2028-29 receipts are £3.1 billion a year lower because the forecast for income from wages and salaries has been revised down by an average of 0.1 per cent each year, largely reflecting weaker expected nominal earnings growth.
• The effective tax rate (ETR) has been revised up by an average 0.2 percentage points each year, raising receipts by an average of £3.4 billion from 2023-24 onwards.^[24]This is largely due to the extra £6.5 billion of PAYE received earlier in 2023-24 recently identified by HMRC,^[25] which offset slightly lower-than-expected PAYE receipts outturn in the months since our November forecast. These upward revisions diminish over the forecast as growing downward revisions to average earnings weaken the impact of fiscal drag on the ETR.
• Measures announced in this Budget reduce revenue by an average £9.7 billion in each year of the forecast, with the bulk from lower NICs (both employee and self-assessed) receipts.
• The indirect effects of labour supply measures, including the 2 percentage point cut to both NICs employee and NICs self-employed main rates, and the changes to the high-income child benefit charge threshold, add an average £0.7 billion of receipts each year, by incentivising individuals to join the labour force (we estimate a total employment increase of 31,000) and those already working to increase their hours (by an equivalent 92,000 in full-time equivalent terms.^[26]

Table 4.2: Non-SA income tax and NICs: changes since November 2023

Self-assessment (SA) income tax

4.15 Self-assessed income tax receipts are expected to reach £42.5 billion in 2023-24 down by 1.0 per cent on last year. Receipts received so far this year are below the November forecast. However, we expect much of the shortfall relating to payments on 2023-24 liabilities to be offset next year when taxpayers make their final payment in January 2025. This explains our forecast of a sharp rise in receipts in 2024-25. This is also driven by a rise in savings income in 2023-24, due to higher interest rates, as well as the previously announced fall in the additional rate threshold from April 2023, which both impact SA receipts received in 2024-25.

4.16 Relative to our November forecast, we have revised receipts up by an average of £1.3 billion a year. Forecast changes add an average £0.4 billion to each year of the forecast, though by diminishing amounts from 2026 onwards. Upward revisions to self-employment income in the early years of the forecast are gradually and partially offset by weaker savings income in the last three years of the forecast due to lower expected Bank Rate.

4.17 Policies announced in this Budget add an average £1.4 billion to receipts from 2025-26 to the end of the forecast. These measures include: the cut to Class 4 NICs rates adds £0.4 billion in SA income tax receipts by 2028-29 (in large part due to it reducing the incentive to incorporate); reforming the non-domicile regime adds an extra £1.3 billion in self-assessed income tax from 2026-27 onwards; and measures to collect HMRC debt add another £0.4 billion from 2025-26 onwards.

VAT

4.18 VAT receipts are expected to increase by 5.3 per cent to £170.7 billion in 2023-24. This increase is slightly weaker than this year’s expected growth in overall nominal consumption of 5.9 per cent. This relative weakness results from the standard-rated share of household expenditure falling this year by 1.6 percentage points to 48.3 per cent due to increased spending on rent, food and energy, driven by higher prices for these components crowding out standard-rated expenditure. As inflation subsides, we expect the standard rated share of household expenditure to increase back to its historical levels, reaching 49.6 per cent in 2028-29. From 2024-25, VAT receipts are forecast to grow by an average of 3.8 per cent each year, and nominal consumption by an average of 3.7 per cent each year.

4.19 Relative to our November 2023 forecast, receipts have been revised down by £2.6 billion (1.5 per cent) in 2023-24, reflecting weaker-than-anticipated outturn year-to-date. In future years, receipts have been revised down by an average of £2.9 billion (1.5 per cent), mostly reflecting this lower starting point.

4.20 Our forecast is sensitive to our projections of the standard-rated share of consumption and the VAT gap. Since 2011, the standard-rated share of household consumption has ranged from 47.9 per cent in 2020 to 50.1 per cent in 2011. From 2024-25 onwards, assuming a 47.9 per cent standard rate share of household consumption in every year would decrease the forecast by approximately £3.1 billion per year, and assuming a 50.1 per cent share would increase the forecast by approximately £3.0 billion per year. As set out in Box 4.1, the VAT gap is a measure of non-compliance, estimating the difference between actual VAT collected and theoretical VAT liability. This has fallen from 14 per cent in 2005-06 to 5.4 per cent in 2021-22 as a proportion of the VAT theoretical liability. We assume it will remain around its current levels over the forecast period, reflecting an expectation of diminishing marginal returns from tax gap reducing activity.

Onshore corporation tax

4.21 Receipts of onshore corporation tax are expected to reach £92.2 billion in 2023-24 and have jumped by almost 25 per cent from last year. The strength in receipts reflects strong profit growth for both financial and non-financial firms (up 11.5 per cent and 5 per cent respectively in 2023) as well as the rise in the main rate of corporation tax from 19 per cent to 25 per cent from April 2023. Receipts growth was particularly strong for financial sector firms where retail banks benefited from the rise in their net interest margins.

4.22 Growth in 2024-25 is expected to slow to 6.7 per cent, more than explained by companies being subject to a full year of the higher rate of corporation tax. In contrast to 2023, profits for both financial and non-financial companies are expected to decline by 0.6 per cent and 1.8 per cent in 2024. The stickiness of wage settlements coupled with lower inflation is likely to squeeze the profit margins of non-financial firms while some moderation in banks’ net interest margins is likely to affect financial sector profits.

4.23 Onshore corporation tax receipts have repeatedly surprised to the upside in recent years. Profits have proved more resilient than expected while the effective tax rate on these profits (abstracting from the rate rise) has risen. The rise in the effective tax rate is broad-based across sectors that have historically been big payers of corporation tax such as financial services, retail and professional services. We have assumed that most of this strength in receipts relative to profits will be maintained through the forecast.

4.24 These factors, coupled with the rise in the main rate of corporation tax (only partly offset by the more generous capital allowances regime) mean that onshore corporation tax is estimated to be 3.4 per cent of GDP in 2023-24, up from 2.5 per cent of GDP in 2020-21. Chart 4.4 shows that we forecast onshore corporation tax then stabilises at 3.5 per cent of GDP from 2024-25 onwards, which would be the highest level since the introduction of corporation tax in 1965.

Chart 4.4: Onshore corporation tax rates and receipts as a share of GDP

4.25 Relative to our November 2023 forecast, onshore corporation tax has been revised down by an average of £2.5 billion a year prior to policy measures announced in this Budget (Table 4.3).

• Lower profits from both financial and non-financial firms take an average of £2.7 billion a year off receipts. Non-financial profits have slowed more rapidly in the second half of 2023 than anticipated and remain weaker over the following two years than in November.
• The direct effects of Budget measures reduce receipts by £0.9 billion by 2028-29 mainly due to the reduced incentive to incorporate from the cuts to NICs. This is partly offset by a small boost to profits from the indirect effects of the Budget measures.

Table 4.3: Onshore corporation tax: changes since November 2023

Oil and gas receipts

4.26 Oil and gas receipts include offshore corporation tax (‘ring fence’ corporation tax and the supplementary charge), petroleum revenue tax (PRT) and the temporary energy profits levy (EPL), which together create a 75 per cent tax rate on North Sea profits until the end of March 2029. We expect these taxes to raise £5.2 billion in 2023-24 – a £4.6 billion fall in receipts on 2022-23 – due to sharply lower energy prices and lower production. We expect receipts to fall further from 2023-24 to reach £2.2 billion by 2028-29 as energy prices and production decline.

4.27 Relative to our November forecast, we have revised receipts down by an average of £1.7 billion between 2024-25 and 2028-29. This is primarily due to the lower path of gas prices. The one-year extension of the EPL to 2028-29 raises the forecast relative to November by £1.2 billion in the final year. The conflict in the Middle East, including disruptions to shipping in the Red Sea, has had little impact on oil and gas prices so far but remains a source of uncertainty for the forecast, as set out in Box 2.2.

Fuel duties

4.28 Fuel duty receipts are expected to raise £24.6 billion this year and remain relatively flat next year as a result of the extension of the 5p cut and the freezing of duty rates announced in this Budget. Receipts are forecast to increase by £2.5 billion (10.3 per cent) in 2025-26, on the assumption that the government implements the reversal of the 5p cut and uprates with RPI, and then to continue rising, albeit at a slowing rate, as assumed higher duty rates are increasingly offset by a reduction in the tax base due to the electrification of the vehicle fleet. We assume that 6 per cent of total car kilometres are driven by electric vehicles in 2023-24, growing to 21 per cent in 2028-29. If the electric vehicle share of car kilometres were one percentage point higher (lower) across the forecast period, that would reduce (increase) fuel duty receipts by £0.2 billion a year, on average.

4.29 Compared to our November forecast, receipts are £0.3 billion higher in 2023-24 due to higher-than-anticipated year-to-date outturn, and are £2.2 billion a year lower, on average, across the rest of the forecast. This downward revision is driven by: lower RPI inflation reducing duty rates (lowering the forecast by £0.7 billion a year); revising down growth in vehicle miles (reducing the forecast by £0.6 billion in 2028-29); and, the one-year extension of the 5p cut and the duty rate freeze announced in this Budget which reduces receipts by £3.1 billion in 2024-25 and by £0.8 billion thereafter.

4.30 Our forecast now assumes that fuel duty will rise by 5p in March 2025 and will be uprated with RPI in every year from April 2025 onwards, in line with stated government policy. In practice, these policies have rarely been implemented: the reversal of the 5p cut has been delayed twice, and fuel duty has not been uprated with RPI since 2011. To illustrate the risk associated with this assumption, Chart 4.5 shows that if the duty rate were to remain unchanged at its current level throughout the forecast period, it would reduce receipts, on average by £3.7 billion between 2025-26 and 2028-29, and by £4.8 billion in 2028-29. In Chapter 5 we show how this affects headroom against the Chancellor’s fiscal targets.

Chart 4.5: Fuel duty: forecasts versus outturn

Other receipts

4.31 Property transaction taxes receipts are expected to raise £12.7 billion in 2023-24, a 23.7 per cent fall from last year due to the downturn in the housing market. We expect receipts to rise in 2024-25 to £14.0 billion as the housing market recovers and reflecting expected forestalling ahead of the increase in nil-rate thresholds that is due to come to an end in March 2025. With housing transactions then forecast to rise steadily, receipts reach £22.1 billion in 2028-29. Relative to our November forecast, receipts are slightly higher across the forecast driven by a stronger outlook for the housing market. We also assume a small near-term increase in transactions due to the Budget policy lowering the higher rate of CGT for residential property.

4.32 Capital gains tax (CGT) receipts are expected to be £14.8 billion in 2023-24, a 12.3 per cent fall from last year. This fall represents a return to the long-run trend in receipts growth after two years of unusually high-value disposals possibly related to forestalling in advance of a rate rise recommended by the Office for Tax Simplification that did not actually occur. We forecast receipts to rise again from next year, following the profile of equity prices, and to be given a further boost by reform of the non-domicile regime which adds an average of £1.3 billion in each of the last three years, with receipts reaching £23.5 billion in 2028-29. The overall impact of policy and forecast changes relative to November raises receipts by an average £0.7 billion per year from 2024-25.

4.33 Inheritance tax (IHT) receipts are expected to raise £7.6 billion this year, an increase of 6.4 per cent on last year. We then expect it to fall slightly before picking back up over the rest of the forecast. This in large part reflects the growth path of house prices in the near term, and then the rise in share prices over the remainder of the forecast. Relative to November, receipts are largely unchanged across the forecast, with slight increases in 2025-26 and 2026-27 mostly due to higher house prices.

4.34 Alcohol duty receipts are expected to raise £12.6 billion this year, up by £0.2 billion from last year, driven by the 10.1 per cent duty rate increase in August 2023, partially offset by lower consumption.^[27] Receipts are expected to grow by an average of 4.9 per cent throughout the rest of the forecast, rising to £16.0 billion in 2028-29, driven by increases in duty rates with RPI and modest increases in alcohol consumption over the forecast. Compared to our November forecast, receipts are down by £0.3 billion in 2023-24, due to lower-than-anticipated outturn, and by an average of £1.0 billion through the rest of the forecast due mainly to a lower RPI forecast and the freeze in duty rates announced at this Budget.

4.35 Tobacco duty receipts are expected to raise £8.8 billion this year, down £0.6 billion, (6.5 per cent) on 2022-23. This is due to a significant fall in consumption, offsetting duty rate increases. Tobacco receipts are projected to fall in each year of the forecast, reaching £8.2 billion in 2028-29. This results from declining forecast tobacco consumption, which more than offsets increases to the duty rate, including the additional duty rate increase in October 2026 announced in this Budget. Relative to November, the forecast is down an average of £0.5 billion a year, mostly driven by a downward revision to our tobacco consumption assumptions.

4.36 Air Passenger Duty (APD) receipts are expected to raise £3.8 billion this year, exceeding their pre-pandemic peak for the first time as the sector continues to recover. Receipts are forecast to increase to £4.5 billion next year, driven by the continued recovery in passenger numbers and then continue to increase, by 7.1 per cent a year on average, driven by increasing passenger numbers and higher duty rates. Relative to our November forecast, receipts have been revised down by £0.1 billion a year, on average.

4.37 Emissions trading scheme (ETS) receipts are expected to raise a record £6.1 billion this year before dropping to £3.6 billion next year and then decreasing through the forecast period, due to a reduction in the number of permits being auctioned, to reach £1.6 billion in 2028-29.^[28] Since November, receipts have been revised down by an average of £0.5 billion a year driven by a lower carbon price. The ETS forecast is highly sensitive to movements in carbon futures prices, which have been very volatile, decreasing from a peak of £109 per tonne of CO2 emissions in August 2022 to a low of £31 in January 2024.^[29]

4.38 Electricity generator levy (EGL) receipts are expected to raise £1.3 billion this year. Estimates have been revised down since November by an average of £0.3 billion due to lower-than-expected wholesale electricity prices. Based on current expected prices, EGL revenues are expected to fall to zero in 2027-28 as the forecast for generators’ selling prices is below the benchmark price.

4.39 Receipts from various environmental levies are expected to raise £9.9 billion this year. These have been revised up by £0.6 billion a year, on average, since our November forecast.^[30] This is mainly driven by the contracts for difference scheme, where lower
wholesale electricity prices are increasing the subsidy paid to generators, which in turn increases the receipts collected from suppliers to fund the subsidy. These levies, excluding the green gas levy, are fully offset in our spending forecast, so are neutral for public sector net borrowing.

4.40 VAT refunds are projected to be £27.6 billion this year. They then grow over the forecast in line with government consumption, reaching £30.9 billion in 2028-29. Relative to November, the forecast has been revised down by an average of £0.8 billion per year, due to a lower forecast of government consumption. Like environmental levies, VAT refunds are offset in spending and are neutral for public sector net borrowing.

4.41 Interest and dividend receipts include income from the Government’s financial assets. These are expected to raise £41.1 billion this year. The sharp rise in receipts in 2023-24 reflects both the effect of higher interest rates on interest-sensitive elements such as government bank deposits and foreign exchange reserves, and the effect of higher RPI inflation on accrued student loan interest. Market expectations indicate that interest rates will start to fall, while RPI inflation has already dropped from its recent peaks, resulting in interest and dividend receipts falling back in 2025-26 and 2026-27. Relative to November, receipts are £0.6 billion higher a year on average from 2024-25 onwards. Modelling changes offset the effect of lower interest rates and RPI inflation.

4.42 Business rates are expected to raise £29.5 billion this year, and to grow by 9 per cent in 2024-25 and then a further 10 per cent in 2025-26. Business rates are calculated by multiplying the rateable value of non-domestic properties by the multiplier (uprated by CPI inflation). In 2024-25, growth in England will be boosted by the 6.7 per cent rise in bills for those properties affected by the standard multiplier, plus less generous transitional relief related to the 2023-24 revaluation. In 2025-26, the £2.5 billion relief for the retail, hospitality and leisure sectors in England is assumed to end with transitional relief again diminishing. Thereafter, growth is expected to average just under 2 per cent a year over the remainder of the forecast. Relative to our November 2023 forecast, business rates are expected to be £0.7 billion a year lower from 2024-25 onwards, primarily due to a lower path for CPI inflation. Budget policy measures have little overall effect as a more generous business rates regime for film studios is offset by reforms to empty property relief.

4.43 Gross operating surplus (GOS) is expected to be £73.5 billion this year. This has been revised up, compared to November, by an average of £3.9 billion a year from 2024-25 to 2028-29. This is driven by higher general government depreciation, which is due to higher ONS outturn on capital stocks increasing depreciation in 2022-23 and feeding through the forecast. Public sector depreciation, which accounts for the majority of GOS, is offset in spending and is neutral for public sector net borrowing.

Public sector expenditure

Summary of the expenditure forecast

4.44 From a post-war peak of 53.1 per cent of GDP in the midst of the pandemic in 2020-21, total public spending falls as a share of the economy to an estimated 44.5 per cent in 2023-24 (Chart 4.6, Chart 4.7 and Table 4.4). We then expect it to continue to fall over the next two years with a decline in debt interest costs more than offsetting an increase in welfare spending. From 2025-26 onwards, the further decline is due to departmental spending (resource and capital departmental expenditure limits or DEL) growing more slowly than the economy, a gradual fall in forecast welfare spending as a share of GDP, and a decline in other elements of annually managed expenditure (AME), including unfunded public service pensions, environmental levies, and student loans.

4.45 As shown in Chart 4.6, spending is now forecast to be 0.2 per cent of GDP lower in 2028-29 than in our November forecast, reflecting a 0.2 percentage points reduction in pre-measures spending, primarily because of lower forecast inflation and interest rates reducing debt interest. The direct and indirect effect of policy measures in this Budget leave spending as a share of GDP broadly unchanged in 2028-29.

Chart 4.6: Public expenditure as a share of GDP

Chart 4.7: Public spending as a share of GDP by category

Table 4.4: Total managed expenditure (TME) split between departmental expenditure limits (DEL) and annually managed expenditure (AME)

4.46 Public spending as a share of GDP in 2028-29 is forecast to be 2.9 per cent of GDP higher than it was in 2019-20, prior to the pandemic (Chart 4.8). That reflects both pre-existing trends and subsequent policy changes:

• The spending plans in place in March 2020 would have resulted in the spending-to-GDP ratio falling by 0.8 per cent of GDP over this period.
• Underlying forecast revisions raise spending by 3.0 per cent of GDP relative to our pre-pandemic expectations. Within this total, debt interest spending contributes 2.0 percentage points, due to upward revisions to both the amount of debt and the cost of servicing it. Revisions to welfare spending contribute a further 1.5 percentage points, due to higher inflation and increases in health- and disability-related benefit caseloads. Other forecast revisions reduce spending by 0.5 per cent of GDP.
• Discretionary spending increases announced between March 2020 and Spring 2022 would have raised the spending-to-GDP ratio by 1.4 percentage points, largely due to departmental spending increases in the March 2020 Budget and the October 2021 Spending Review.
• Discretionary spending reductions announced between the 2022 Autumn Statement and March 2024 lowered spending by 0.7 per cent of GDP in 2028-29. This is more than explained by reductions to departmental spending in Autumn 2022, partially offset by the more generous provision of free childcare hours and higher defence spending in March 2023.

Chart 4.8: The rise in the spending-to-GDP ratio between 2019-20 and 2028-29

Changes in spending since our November 2023 forecast

4.47 Relative to our November forecast, spending in cash terms has been revised down in each year of the forecast (Table 4.5). This is more than explained by underlying forecast differences, primarily lower debt interest spending caused by lower interest rates and inflation. This is slightly offset by higher PSNB-neutral spending, notably higher depreciation, which is offset exactly by higher receipts.

4.48 Policies in this March Budget raise overall spending by an average of £0.3 billion a year over the forecast. This is mainly driven by higher capital spending on the government’s public sector productivity programme focussed on the NHS (£0.9 billion over three years), and the additional spending from raising the high-income child benefit charge (HICBC) threshold. These are partly offset by decreases in resource DEL spending which average £0.7 billion per year. The indirect effects of the policy package raise spending by a further £0.6 billion a year on average, due to the costs of servicing the additional debt issued to finance the overall fiscal loosening.

Table 4.5: Public sector expenditure: changes since November 2023

Spending within departmental expenditure limits

4.49 Spending subject to departmental expenditure limits (DELs) makes up just over 40 per cent of all public spending. In this section, ‘RDEL spending’ refers to departmental resource, or day-to-day, spending, and ‘CDEL spending’ refers to departmental capital, or investment, spending.^[31] Our latest forecast (Tables 4.7 and 4.8) reflects:

• Departments’ detailed departmental plans for 2023-24 and 2024-25, as announced in the October 2021 Spending Review, incorporating the latest 2023-24 supplementary estimates as set out by the Treasury in February 2024^[32], and the effects of policy announcements since then. In these years, the Government has set department-by-department plans in cash terms and put in place tight spending controls which are generally respected.
• The Government’s post-Spending Review spending assumptions for 2025-26 and beyond, which set two overall spending totals (one for total resource DEL and one for total capital DEL) but provide no detailed plans for how these are to be divided between each government department. As we approach the end of the period covered by the 2021 Spending Review, this overall spending envelope covers four of the five years of our forecast. As discussed in Box 4.2, these indicative spending totals have, historically, been subject to significant upward revisions prior to or at the conclusion of the Spending Review process which allocates these envelopes out between departments.

4.50 Compared to our November forecast, departmental resource spending is on average £4.8 billion lower over the forecast period. This reflects a reclassification of some business rates relief measures as AME (which is offset in higher AME spending) and a cash decrease to the Government’s post-Spending Review RDEL plans due to lower inflation from 2025-26 onward. Departmental capital spending is on average £0.6 billion higher over the forecast largely reflecting the government’s public sector productivity programme focused on the NHS. Both are discussed in further detail below.

Chart 4.9: Breakdown of total departmental spending in the forecast

Departmental spending in 2023-24 and 2024-25

4.51 Departmental resource spending in 2023-24 is expected to be £422.9 billion, up £15.5 billion from 2022-23, and is forecast to rise to £430.2 billion in 2024-25. The main changes compared to our November forecast are an increased allocation by Treasury of £3.2 billion in 2023-24 at supplementary estimates (partly a switch from CDEL), and our judgement that there will be a higher resource underspend (£1.8 billion in 2023-24) based on the latest in-year spending data (Table 4.6). A further change is the reclassification of a subset of business rate reliefs (£5.0 billion in 2023-24 and £5.1 billion in 2024-25) from DEL to annually-managed expenditure. This change aligns the treatment of these with the normal treatment of tax reliefs in the public spending framework. Compared to our November forecast, departmental resource spending is £3.6 billion lower in 2023-24 and £5.3 billion lower in 2024-25.

4.52 Departmental capital spending in 2023-24 is expected to be £95.6 billion, up £5.8 billion on 2022-23, and is forecast to rise to £99.2 billion, in 2024-25. The main changes compared to November forecast are a £3.6 billion in-year capital reduction by Treasury at supplementary estimates (partly a switch to RDEL), and judgement of a lower capital underspend (£3.1 billion in 2023-24) based on the latest in-year spending data (Table 4.7). Compared to our November forecast, departmental capital spending is £0.4 billion lower in 2023-24 and unchanged in 2024-25.

Departmental spending from 2025-26 onwards

4.53 From 2025-26 onwards, we use the Treasury’s stated assumptions about the overall RDEL and CDEL envelopes. At this Spring Budget these assumptions are that:

• RDEL grows by 1.0 per cent in real terms on average from 2025-26. In cash terms this means RDEL spending compared to November is £0.8 billion lower on average from 2025-26, due to lower forecast inflation and the impact of some small Budget policies.^[33] In addition, the reclassification of some business rates reliefs to AME leads to a further £3.9 billion reduction, which is broadly neutral for overall public spending. Overall, this means that cash resource DEL spending is on average £4.7 billion lower per year between 2025-26 and 2028-29 compared to our November forecast.
• CDEL is held broadly flat in cash terms from 2025-26. Compared to November, CDEL spending is expected to be £0.8 billion higher between 2025-26 and 2027-28, almost all of which reflects an allocation for the government’s public sector productivity programme.^[33] It is unchanged in cash terms in 2028-29.

Table 4.6: Departmental resource spending: changes since November 2023

Table 4.7: Departmental capital spending: changes since November 2023

4.54 The real per person value of departmental spending plans has fallen significantly since the Government first set them out in the October 2021 Spending Review and is now forecast to be flat over the next five years. In current prices, total departmental spending^[34] in the forecast accompanying the 2021 Spending Review was initially expected to be £8,060 per person in 2021-22, before falling to £7,920 in 2023-24, and the rising to £8,210 in 2026-27, based on our economic forecast at that time. However, both inflation and population growth have subsequently turned out to be higher than we forecast at that time. As a result, and on the basis of the Government’s latest spending assumptions, real departmental spending in this forecast is expected to be £7,580 per person in 2026-27. This is around £630 (8 per cent) lower than anticipated when plans were first set in October 2021.^[35]

4.55 Chart 4.10 shows how the real per person value of these departmental spending has evolved over time. The initial October 2021 SR plans (blue line) implied a £150 (2 per cent) increase in real per person spending by 2026-27. The roughly £630 (8 per cent) erosion in real per person spending power since then has been driven by:

• DEL policies announced at fiscal events between October 2021 and March 2024 (yellow line) which increased DEL spending in cash terms between 2022-23 and 2025-26, but left DEL spending in 2026-27 largely unchanged;
• higher than expected inflation (green line) in the interim which reduces forecast real per person spending by £470 (6 per cent) in 2026-27;
• higher population growth (purple line) which reduces forecast real per person spending by a further £160 (2 per cent) in 2026-27; and
• as a result, in our latest forecast, real per person spending on public services no longer grows over the next five years.

Chart 4.10: Real departmental spending per person (2023-24 prices)

Box 4.2: The Government’s post-Spending Review departmental spending plans

The plans for the Departmental Expenditure Limits (DELs) which the Government set in the October 2021 Spending Review (SR21) run to the end of 2024-25. This means that, for the over 40 per cent of public spending within DEL, the Government currently only has detailed, department-by-department spending plans covering one year of our five-year fiscal forecast. This shrinking of the departmental financial planning horizon down to one year or less has been a feature of the UK’s public spending framework since 2007.

Chart B: Months covered by departmental spending plans

In the absence of detailed departmental spending plans after 2024-25, departmental spending follows two overall envelopes set by the Government: that day-to-day spending (RDEL) grows by an average of 1.0 per cent a year in real terms; while capital spending (CDEL) is broadly flat in cash terms. Taken together, these assumptions imply total DEL falls as a share of GDP by 1.0 percentage points over the final four years of the forecast, though in 2028-29 DEL is still 1.1 percentage points above its pre-pandemic share of GDP).

While these spending envelopes are not yet allocated to departments, we can still explore their implications by considering what existing input targets and commitments in some areas of spending imply for growth in spending in areas not covered by such targets – often called ‘unprotected’ spending. Looking at day-to-day departmental spending, we assume (Chart C):

• Spending on the NHS in England grows by 3.6 per cent a year in real terms, using the IFS’ central scenario for the Government’s Long-Term NHS Workforce Plan^a. This is in line with the long-run (1949–50 to 2022–23) average real terms growth rate of UK health spending, though above real terms growth in recent years. The government’s public sector productivity package could increase NHS productivity and partially offset the spending growth required to deliver the NHS Workforce Plan
• Defence spending is held flat as a share of GDP, consistent with the Government’s commitment to keep such spending above the NATO minimum of 2 per cent of GDP. Meeting the Government’s ambition to increase defence spending to 2.5 per cent of GDP would increase the pressure on unprotected spending.
• Spending on Official Development Assistance (ODA) is maintained at 0.5 per cent of gross national income (GNI) throughout the forecast. If the Government were to return ODA spending to 0.7 per cent of GNI – as they have committed to when the fiscal situation allows^b – this would also increase the pressure on unprotected spending.
• Core schools spending is held flat in per-pupil real terms, reflecting the Government’s statement that SR21restored per pupil spending to 2010 levels in real terms.^c
• The additional £3.9 billion per year announced in the March 2023 Budget to fund the extension of 30 hours of free childcare to parents of nine-month- to two-year-olds is protected in cash terms.
• The consequences of our NHS and schools spending assumptions for devolved administrations are captured using the Barnett formula.

With the post-SR21 envelope for total RDEL spending provided by the Treasury, these assumptions would leave other ‘unprotected’ RDEL spending (accounting for just under a third of day-to-day departmental spending) needing to fall by 2.3 per cent a year in real terms from 2025-26. If defence and ODA spending increased in line with the Government’s ambitions outlined above, this would lead to unprotected spending needing to fall by an average of 3.6 per cent a year.^d Delivering a 2.3 per cent a year real terms fall in day-to-day spending would present challenges. Performance indicators for public services continue to show signs of strain^e and there are other near-term fiscal risks, such as the potential need for countries to provide a greater amount of aid to Ukraine and alleviate funding pressures on local authorities. Longer-term pressures on public spending, such as from climate change and an ageing population, are also building, as discussed in our Fiscal risks and sustainability reports.

Chart C: Implied post-SR21 breakdown of real RDEL spending and share of RDEL

The implementation of these assumptions for departmental spending therefore remain a significant risk to our forecast. As Spending Reviews have approached, governments since 1998 have increased the annual real growth in day-to-day spending by an average of 1.1 per cent. In the run-up to the two Spending Reviews in November 2015 and October 2021, governments topped up the RDEL envelope by an average of £39 billion and £32 billion per year respectively. Such increases have been a major cause of the differences between our past borrowing forecasts and outturn, explaining around half of the difference on average between 2011-12 and 2021-22.^f

Chart D: Change in real RDEL envelope growth rate at Spending Reviews

a)Institute for Fiscal Studies, Implications of the NHS workforce plan, August 2023.

b) Specifically, ‘when the independent Office for Budget Responsibility’s fiscal forecast says that, on a sustainable basis, the UK is not borrowing to finance day-to-day spending and underlying debt is falling’.

c) Autumn Budget and Spending Review 2021 speech.

d) These ambitions are for total DEL, so to calculate the RDEL implications we assume the increase is in proportion to the 2024-25 split between RDEL and CDEL.

e) Institute for Government, Performance Tracker 2023, October 2023.

f) Half of the three year ahead forecast difference. See: Atkins, G and Lanskey, L, Working paper No.19: The OBR’s forecast performance, August 2023.

Welfare spending

4.56 Total welfare spending in our forecast refers to AME spending on social security and tax credits. Around half is subject to the Government’s ‘welfare cap’, which excludes the state pension and those payments most sensitive to the economic cycle (we discuss performance against the cap in Chapter 5). Our welfare spending forecasts are based on the determinants in our latest economy forecast – principally population, unemployment, earnings, and inflation – and informed by the latest outturn data and Department for Work and Pensions models.

4.57 Welfare spending is forecast to rise sharply this year (by £34.4 billion, or 13.1 per cent) and next (by £19.3 billion, or 6.5 per cent), driven by the uprating of most benefits with CPI inflation. Benefits were uprated by 10.1 per cent in April 2023 and are expected to be uprated in April 2024 by 6.7 per cent, the September 2023 CPI inflation figure. Welfare spending is then forecast to rise by an average of £11.2 billion (3.4 per cent) a year over the rest of the forecast period. The main drivers of this increase are higher pensioner spending due to the ageing population and the triple lock, and rising caseloads for health and disability benefits. Spending on these components explains more than nine-tenths of the increase in total welfare spending between 2024-25 and 2028-29.

Table 4.8: Total welfare spending

Chart 4.11: Welfare spending as a share of GDP

4.58 Relative to our November 2023 forecast, welfare spending is now expected to be slightly lower in all years by an average of £2.8 billion. Table 4.9 shows that this revision is driven by:

• Lower uprating due to falls in our inflation and earnings forecasts (averaging £2.9 billion lower from 2025-26). As inflation is not binding on the triple lock for any forecast years, the majority of downwards pressure from lower inflation falls on child and working-age benefits.
• Lower spending on the in-work component of universal credit (averaging £2.6 billion lower each year). We have revised down our caseload forecast due to a larger-than-expected proportion of tax credits claimants closing their claim rather than migrating to UC. In addition, higher-than-expected earnings growth among UC claimants has led to larger earnings tapering of awards.
• Other changes which include: higher state pension and disability caseloads (averaging £1.0 billion a year), mainly due to lower mortality assumptions in the state pension forecast and an assumption that a higher proportion of personal independence payment claims will be awarded, and higher fraud and error costs in universal credit (averaging £0.6 billion a year), mainly due to higher-than-expected fraud in current claims.
• Policy measures in the Spring budget which increase spending by an average of £0.1 billion over the forecast. The largest cost is additional spending on child benefit from raising the threshold beyond which the high-income child benefit charge (HICBC) applies, which costs £0.4 billion over the forecast. This is partially offset by reductions in spending on Universal Credit due to claimants working more hours as a result of the two-percentage point NICs cut (£0.2 billion), and reductions in spending on PIP from additional funding to support the processing of increased volumes of disability benefit claims (£0.1 billion).

Table 4.9: Welfare spending: changes since November 2023

4.59 Volatility in our inflation and earnings forecasts, given their impact on uprating, is a key cause of uncertainty for our welfare forecasts. Over our previous four forecasts, the average annual change in welfare spending due to changes in uprating has been £8.4 billion. A 1 percentage point increase in our inflation and earnings forecasts next year would increase uprating costs by £3.1 billion from 2025-26.

Locally financed expenditure and public corporations’ expenditure

4.60 We expect locally financed current expenditure to rise from £62.1 billion in 2023-24 to £75.4 billion in 2028-29, as local sources of income grow steadily by an average of £2.7 billion (4.3 per cent) a year.^[36] Compared with the November 2023 forecast, locally financed current expenditure is broadly unchanged (Table 4.10):

• Net use of current reserves outturn has been revised down by £0.3 billion to £1.2 billion in 2023-24. Due to the ongoing funding pressures on local authorities, we assume that there will be further drawdowns during the current Spending Review period of £0.6 billion in 2024-25.
• Additional council tax flexibilities for Birmingham, Thurrock, Slough and Woking announced in the local government settlements for 2024-25 and an increase in referendum principles for Police and Crime Commissioners authorities have increased forecast council tax receipts by an average of £0.1 billion a year.

4.61 We expect locally financed capital expenditure to fall from £10.8 billion in 2022-23 to £8.2 billion in 2028-29.^[37]
This is driven by lower local authority borrowing for capital expenditure which is forecast to fall from its 2019-20 peak of £12.2 billion to £6.6 billion in 2028-29. This reflects higher interest rates on loans from the Public Works Loan Board, local authorities’ principal source of financing, and new restrictions on commercial borrowing. Public corporations’ capital expenditure is expected to grow from £11.2 billion in 2023-24 to £11.7 billion in 2028-29. These forecasts are broadly unchanged from November.

4.62 The wider pressures on local authority finances are a risk to our forecast. Since our last forecast, Nottingham City Council has issued a ‘section 114’ notice, adding to the eleven notices already issued since 2018.^[38] These notices indicate that the authority’s forecast income is insufficient to meet its forecast expenditure for the next year. Since 2010-11, local authority spending has fallen from 7.4 to 5.1 per cent of GDP in 2022-23, and it falls further in our forecast to 4.7 per cent of GDP in 2028-29. Given local authorities’ statutory duty to provide a range of services where demand is likely to continue to grow, for example adult and child social care, pressure on local authority finances and services will continue.

Table 4.10: Locally financed and public corporations’ expenditure: changes since November 2023

Debt interest spending

4.63 From a post-war high of 4.4 per cent of GDP in 2022-23, debt interest spending is forecast to fall to 3.1 per cent of GDP by 2025-26 due mainly to the forecast decline in RPI inflation reducing the cost of index-linked gilts. It is then forecast to increase to 3.4 per cent of GDP, or 8.3 per cent of total revenue in 2028-29 Chart 4.12, due mainly to the rising stock of debt.

4.64 Compared to our November forecast, debt interest is expected to be on average 0.4 percentage points lower as a share of GDP and 1.0 percentage points lower as a share of total revenue between 2023-24 and 2028-29. This reflects a lower forecast for RPI inflation, Bank Rate and gilt yields. Box 4.3 details the uncertainties around our debt interest spending forecast and its sensitivity to market determinants in further detail.

Chart 4.12: Debt interest spending relative to GDP and revenues

Box 4.3: The sensitivity and volatility of debt interest spending

In the decade prior to the pandemic, debt interest costs were remarkably stable at around £40 billion per year. Interest costs were often a source of ‘upside’ surprises to the fiscal outlook – as outturn interest rates typically turned out to be lower than our forecasts derived from market expectations (Chart E). As a result of this and the fiscal savings from the Bank of England’s quantitative easing (QE) operations (which swapped relatively higher yielding gilts for lower yielding bank reserves), actual debt interest costs were around £27 billion on average below our five-year-ahead forecast between our June 2010 and March 2016 forecasts.

Chart E: Successive OBR debt interest spending forecasts

The rapid rise in global interest rates over the last two years means that interest costs have soared and market expectations of future interest rates have become more volatile, with large revisions, both upwards and downwards, between forecasts. With Bank Rate and gilt yields spiking to over 15-year highs and RPI inflation reaching a 40-year high in 2022-23, debt interest spending tripled in cash terms to over £100 billion and reached a post-war record of 10.8 per cent of government revenue. This increase in debt interest costs was exacerbated by the higher stock of debt, its shorter maturity, and the prevalence of index-linked debt in the UK’s public liabilities.

The stock of debt is forecast to remain above 90 per cent of GDP which, when combined with the potential for continued volatility in interest rate expectations, means that interest spending is likely to remain a major source of risk to the fiscal outlook. By way of illustration, our forecast for debt interest spending in 2027-28 increased by £18.7 billion when interest rate expectations rose between March and November 2023, but has fallen back by £12.8 billion in this forecast as those expectations moderated. In both cases, this volatility in interest costs has been more than the headroom the Chancellor has set aside against his fiscal mandate.

Previous EFOs and Fiscal risks and sustainability reports have explored how the Bank’s QE operations have shortened the effective maturity of the consolidated liabilities of the public sector.^a Since then, the steady unwinding of the Bank’s gilt holdings via quantitative tightening has begun to reverse some of that effect. As shown in Chart F, the unwinding of 15 per cent of the Bank’s peak gilt holdings has extended the median maturity of the UK’s consolidated public liabilities from 1 year in 2021 to 3 years by the end of 2023. By 2028, when we assume 71 per cent of gilts will have been returned to the market, the effective maturity will have reached 7 years. This is the same as its pre-QE level, reflecting the relatively long effective maturity of gilts, Treasury bills, and National Savings and Investment products whose combined maturity has remained relatively stable between 7 and 9 years since 2010. This steady re-extension of maturities will help to partially re-insulate the public finances from future changes in borrowing costs. But both the UK’s higher debt stock and the potential for continued volatility in interest rates are likely to continue to drive large changes in debt servicing costs from one forecast to the next.

Chart F: Changes in maturity of UK debt stock

a) See our 2023 Fiscal risks and sustainability report and 2021 Fiscal risks report.

4.65 In nominal terms, debt interest spending falls from £111.5 billion in 2022-23 to £88.9 billion in 2025-26, before rising year-on-year to £109.6 billion in 2028-29 (Table 4.11). It has been revised down in each year from 2023-24 relative to November by an average of £13.2 billion, reflecting several factors:

• In the near-term RPI inflation drives the downward revisions, reducing spending by an average of £9.6 billion in the first two years. The RPI impact then falls to an average of £1.2 billion in the final four years as inflation converges towards our November forecast.
• Lower Bank Rate in every year from 2023-24, by an average of 0.7 percentage points, which decreases spending by an average of £6.0 billion a year from 2024-25 onwards (with around two-thirds of this relating to the APF).
• Lower gilt yields, by an average of 0.5 percentage points a year, which reduce spending by increasing amounts in each year and account for an increasing share of the downward revisions as the effects of lower inflation and Bank Rate subside.
• A lower net financing requirement, which reduces spending by an average of £1.0 billion and by a peak of £1.4 billion at the forecast horizon.

Table 4.11: Central government debt interest (net of APF): changes since November 2023

Other annually managed expenditure

4.66 The main changes to other AME spending since our November forecast include:

• The net cost of unfunded public service pensions has been revised down by an average of £0.2 billion a year, reflecting higher scheme receipts due to an increase in earnings, and reduced scheme expenditure due to lower inflation as pensions entitlements are indexed to CPI. The net cost of unfunded public sector pensions is forecast to fall from £4.6 billion in 2023-24 to £0.6 billion in 2028-29, as scheme receipts grow more quickly than scheme expenditure on average.
• Scottish Government current expenditure has increased by an average of £0.4 billion a year over the forecast, due to a higher Scottish block grant and tax receipts.
• National lottery grants have been revised down by £0.2 billion on average a year. This is driven by lower outturn data from national lottery distributors.
• Company tax credits have been revised down by £0.2 billion on average per year. This is driven by a combination of lower outturn spending that offsets the increased forecast of investment.

Deficit aggregates

Borrowing

4.67 Having peaked at a post-war high of £314.7 billion (15.1 per cent of GDP) at the height of the pandemic in 2020-21, public sector net borrowing is expected to be £114.1 billion (4.2 per cent of GDP) this year (Chart 4.13). It is then forecast to fall to £39.4 billion (1.2 per cent of GDP) by 2028-29. This is slightly higher at the forecast horizon than our November forecast and would be the lowest level of borrowing as a share of GDP since 2001-02 (0.5 per cent of GDP).

Chart 4.13: Public sector net borrowing

4.68 Chart 4.14 decomposes the projected 2.9 percentage point decline in borrowing as a share of GDP between 2023-24 and 2028-29. Around a quarter of the reduction in borrowing is through a rise in the tax-to-GDP ratio, the majority due to rising income tax driven by nominal earnings growth combined with frozen tax thresholds. The remaining three-quarters of the fall in borrowing as a share of GDP is through reduced spending. The largest contribution comes from departmental spending falling by 1.0 per cent of GDP, with roughly equal parts coming from current and capital spending. Another 0.4 percentage points come from debt interest falling as a share of GDP as inflation and interest rates moderate. These downward contributions are partly offset by increased welfare spending, which rises by 0.4 per cent of GDP over the forecast.

Chart 4.14: Contributions to the fall in borrowing relative to 2023-24

4.69 Borrowing in cash terms is lower than our November forecast in 2023-24 by £9.8 billion but then slightly higher in every year of the five-year forecast period, by an average of £1.9 billion (Chart 4.15 and Table 4.12). Overall borrowing between 2023-24 and 2028-29 is broadly unchanged (£0.3 billion lower) compared to the November forecast. This reflects the following changes:

• Pre-measures spending is lower by £10.9 billion in 2023-24 and an average of £16.3 billion thereafter. This is driven largely by an average £14.4 billion reduction in debt interest spending due to lower Bank Rate, gilt yields and near-term RPI inflation. Welfare spending is also lower by an average of £2.7 billion from 2024-25, largely due to lower uprating through a fall in near-term CPI.
• Pre-measures tax receipts are lower by £0.8 billion in 2023-24 and by amounts rising to £14.8 billion in 2028-29. This is largely driven by a reduced forecast for inflation and nominal earnings, which over the medium term drives lower income tax and VAT, and a weaker profits forecast which reduces corporation tax receipts.
• The direct effect of policy decisions increase borrowing by £13.9 billion in 2024-25 falling to £5.9 billion by 2028-29. The cut to NICs rates costs an average of £10.5 billion from 2024-25, and the fuel duty freeze costs £3.1 billion in 2024-25 and an average of £0.8 billion thereafter. These costs are partly offset by several revenue-raising measures – including the changes to the non-domicile regime, HMRC compliance measures, and the extension of the Energy Profits Levy – the yield from which builds over the forecast period. Spending policy decisions have a much smaller net impact on borrowing over the forecast.
• The indirect effect of policy decisions reduce borrowing in most years, largely due to additional revenue generated by the temporary boost to demand from the overall fiscal loosening due to the measures, and by the labour supply benefits of the NICs cut and the HICBC change. This more than offsets the cumulative increase in debt interest spending which results from the measures announced at this Budget.

Chart 4.15: Public sector net borrowing: changes since November 2023

Table 4.12: Public sector net borrowing: changes since November 2023

Other deficit aggregates

4.70 Beyond headline PSNB, several other deficit aggregates provide insights into the state of the public finances. The primary deficit, which excludes net interest spending, is a useful proxy of the extent to which discretionary spending is covered by revenues and is sometimes referred to as a measure of ‘fiscal effort’. It is one of the key drivers of the path of debt over the medium-term, as is explored further in Chapter 5 and Box 5.1. The current deficit, which excludes net investment spending, is a useful proxy for the extent to which spending that mostly benefits today’s population is met by the taxes they pay. And all measures of the deficit can be presented in cyclically adjusted terms, correcting for an estimate of the position in the economic cycle, which provides a rough indication of the underlying or structural deficit.

4.71 In this forecast, these alternative measures of the deficit show that:

• The primary deficit is 1.2 per cent of GDP this year and remains in deficit until 2026-27 where it moves to surplus. By the forecast horizon it reaches a surplus of 1.6 per cent of GDP, which would be the largest primary surplus since 2000-01.
• The current deficit is estimated to fall from 1.7 per cent of GDP this year and reaches a surplus of 0.2 per cent of GDP in 2027-28, rising to 0.4 per cent of GDP by 2028-29. This would be the largest current surplus since 2001-02.
• Cyclically adjusted measures of the deficit are similar to unadjusted metrics in 2023-24 but marginally lower in each subsequent year until 2028-29. This reflects the estimated path of the output gap which initially widens slightly, as spare capacity in the economy increases to 2024-25 but diminishes thereafter and closes by 2028-29.

Financial transactions

4.72 To calculate changes in public sector net debt we combine changes in borrowing with changes in financial transactions and valuation effects. Financial transactions capture the effects of public sector net lending and convert the accrued measures of tax and spending which underpin our forecast for borrowing into the cash flows relevant to debt.^[39] Valuation effects capture changes in the value of the assets or liabilities held by the public sector which count toward the measure of net debt.

4.73 Public sector net debt excluding the Bank of England (PSND ex BoE) rises in cash terms in each year from 2023-24 onwards. The top panel of Table 4.13 breaks down these year-on-year changes in PSND ex BoE into changes in public sector net borrowing, net borrowing by the Bank of England, financial transactions and valuation effects. It shows:

• Public sector net borrowing increases debt by decreasing amounts in each year. In 2023-24 borrowing adds £114.1 billion and accounts for three-fifths of the change in debt, falling to £39.4 billion (two-fifths) of the increase in debt in 2028-29.
• Removing net borrowing by the Bank of England (BoE) adds to PSND ex BoE in each year, by amounts falling to £8.8 billion in 2025-26, before rising back to £20.1 billion in 2028-29. This reflects APF sales and redemption losses as the gilts held within it are sold.^[40]
• Student loans push up on debt by increasing amounts in each year, adding £12.8 billion by 2028-29 – largely thanks to higher student numbers and an increase in the maximum loan available per student in Scotland.
• Cash flow timing effects add £30.6 billion to debt in 2023-24 and an average of £17.9 billion thereafter. This reflects adjustments converting accrued receipts into cash and modified interest on student loans.
• Other financial transactions increase debt by an average of £4.1 billion a year, and valuation effects add £19.6 billion to debt in 2023-24, but lower amounts thereafter until 2028-29. This profile is driven almost entirely by the discount (negative premia) on new gilt issuance, which adds £17.5 billion to debt in 2023-24 but falls away sharply as we assume over time the Debt Management Office (DMO) returns to issuing new gilts at nearer to face value.^[41]

4.74 Table 4.13 shows that, relative to November, PSND ex BoE is forecast to be £12.6 billion lower in 2023-24 but higher by more modest amounts in most years thereafter. The downward revision in 2023-24 reflects lower borrowing alongside a lower contribution to debt from cash flow timing effects. The small revisions subsequently are driven primarily by higher borrowing which is partly offset in each year by lower BoE net borrowing, due to lower losses on APF sales and redemptions reflecting higher gilt prices.

Table 4.13: Drivers of changes in PSND ex BoE

4.75 Total additions from APF-related flows to PSND ex BoE have fallen by an average of £5.6 billion per year compared to our November forecast. Around three-fifths of this is through the reduced interest losses that impact PSNB due to the lower forecast for Bank Rate compared to our November forecast. The remaining two-fifths is through reduced losses as gilts held in the APF are sold or redeemed. Box 4.4 discusses how these changes have impacted our estimate of the lifetime cost of the APF and presents scenarios illustrating its sensitivity to movements in interest and gilt rates.

Box 4.4: The sensitivity of the Asset Purchase Facility to market conditions

Since Bank Rate and gilt yields rose from their record lows in the second half of 2022, the Bank of England’s Asset Purchase Facility (APF) has gone from making a profit to making a loss. Having transferred £123.9 billion of cash profits to the Treasury between January 2013 and October 2022, a total of £49.4 billion has been transferred from the Treasury to cover losses incurred by the APF since then. Our latest estimate of the lifetime cost of the APF is a net loss of £104.2 billion. This is £21.9 billion less than we forecast in November due to Bank Rate and gilt yields being 0.8 and 0.6 percentage points lower from 2024-25 onwards, respectively.

This reduction underscores the sensitivity of this calculation to market conditions (in Box 4.3 we discuss the wider volatility of the debt interest forecast). Alongside our central estimate, in Chart G we present upside and downside scenarios that assume Bank Rate and gilt yields are 100 basis points higher and lower from 2024-25:

• With lower interest rates the lifetime cost is £46.6 billion, which is £57.6 billion lower than the central scenario, comprising a £28.1 billion reduction in interest losses and £29.6 billion reduction in valuation losses.
• With higher interest rates the lifetime cost is £156.9 billion, which is £52.7 billion higher than the central scenario, comprising a £28.0 billion increase in interest losses and £24.7 billion in valuation losses.

Chart G: Forecast of quarterly and cumulative flows to and from the APF

This is not a comprehensive assessment of the overall fiscal impact of the Quantitative Easing programme, which supported the economy, asset prices, and financial markets at various points of stress over the past 15 years. The wider economic and fiscal benefits of these interventions would need to be considered in any comprehensive assessment of the impact of QE.

Financing requirement

4.76 The central government net cash requirement (the amount of cash the government needs to raise each year), is forecast to fall from £149.2 billion in 2023-24 to £97.2 billion by 2028-29. Over the five-year forecast period, gross gilt issuances averages 8.0 per cent of GDP and net issuance averages 3.4 per cent of GDP, which are similar levels to our November forecast (Chart 4.16). However, the cash requirement remains elevated by historical standards, which, alongside the unwinding of APF gilt holdings by the Bank of England, means the private sector needs to absorb relatively high volumes of debt for a sustained period. The forecast for the change in private sector holdings peaks at 8.2 per cent of GDP in 2024-25 and averages 6.3 per cent of GDP over the forecast period, compared to 2.7 per cent of GDP between 2000-01 and 2022-23.

Chart 4.16: UK gilt issuance and change in private sector holdings of gilts

Debt and other balance sheet aggregates

4.77 Public sector net debt excluding the Bank of England (PSND ex BoE) rises from 88.8 per cent of GDP in 2023-24 to 93.2 per cent in 2027-28, before falling to 92.9 per cent in 2028-29 (Chart 4.17). Table 4.13 above sets out the drivers of the year-to-year changes in debt. Relative to our November forecast, debt is 0.3 per cent of GDP lower in 2023-24, reflecting lower borrowing, and then as a share of GDP follows a similar path to the November forecast over the remainder of the period.

4.78 In cash terms debt is slightly lower than in November in each year, by an average of £9.9 billion (Table 4.14). This reflects the downward revisions to our pre-measures forecast for borrowing, which cumulatively decrease debt by £40.4 billion, in combination with financial transactions and valuation effects which reduce debt by a total of £5.6 billion. These are partly offset by the net cost of the Budget policy package which adds £40.4 billion to debt.

Chart 4.17: Public sector net debt (excluding Bank of England)

Table 4.14: Public sector net debt (excluding Bank of England): changes since November 2023

4.79 Headline PSND peaks in 2024-25 at 98.8 per cent of GDP before falling to 94.3 per cent of GDP by the forecast horizon (Chart 4.18). The main differences between PSND and PSND ex BoE reflect the Term Funding Scheme (TFS) and valuation effects on gilts held in the APF.^[42]
These differences reduce over the forecast period as TFS repayments are made and the APF unwinds, with the overall difference narrowing to 1.4 percentage points by 2028-29.

4.80 Chart 4.18 also shows forecasts for public sector net financial liabilities (PSNFL) and public sector net worth (PSNW), which provide wider measures of the balance sheet. These measures both peak in 2024-25 and then fall through to the forecast horizon. While underlying debt rises by 4.2 per cent of GDP between 2023-24 and 2028-29, PSNFL falls by 4.2 per cent of GDP. This reflects the cost of acquiring assets such as student loans which add to debt, whereas both the assets and liabilities are recorded in PSNFL so largely net off. (Inverted) PSNW falls faster still, by 5.6 per cent of GDP over the forecast. In addition to the 4.2 per cent of GDP fall in PSNFL, this reflects a 0.6 per cent of GDP increase in non-financial assets and a 0.5 per cent of GDP fall in unfunded pension liabilities.

Chart 4.18: Four measures of the public sector balance sheet

Risks and uncertainties

4.81 Over recent years, large shocks and their aftermath have often resulted in significant revisions to our economic and fiscal forecasts from one fiscal event to the next. We therefore continue to emphasise the uncertainties around our forecast in the light of rapidly changing economic conditions and the possibility that any of our key judgements could prove significantly too optimistic or pessimistic.

4.82 Key economic uncertainties in this forecast include:

• Changes to Bank Rate and gilt yields since our last forecast have largely been favourable for the public finances, reducing debt interest spending in every year. This is a partial reversal of the large increases in debt interest costs in our November forecast. Box 4.3 looks in more detail at the implications for the public finances of volatility in interest rates. A 1 percentage point increase in Bank Rate and gilt yields across the forecast would increase debt interest spending by an estimated £14.5 billion in 2028-29, and increase losses on the Bank of England Asset Purchase Facility (APF) by £52.7 billion (see Box 4.4).
• Productivity growth, as the key driver of economic growth over the medium and long-term, is central to the economic projections that underpin our fiscal forecast. While we forecast higher productivity growth over our forecast (average of 0.9 per cent) than seen over much of the period post-financial crisis (around 0.5 per cent), it is still below the pre-crisis trend of 2 per cent. Scenario analysis in our November 2023 EFO estimated that if productivity was 0.5 percentage points above or below our central forecast, it would reduce borrowing by an estimated £46.0 billion or increase it by £42.2 billion, respectively, by the end of our forecast.
• The prospect of widening conflict in the Middle East poses a significant risk to the path of energy prices as well as supply chain resilience. In Box 2.2 and Chapter 5, we set out a downside scenario where a widening of the conflict causes quarterly inflation to spike to around 7 per cent and also push interest rates higher. We estimate this could raise borrowing, on average, by £23.1 billion a year and leave underlying debt 0.8 per cent higher as a share of GDP by the forecast horizon.
• A major change in our forecast since November has been incorporating updated net migration projections and LFS data from the ONS. Future migration levels are highly uncertain and difficult to forecast, particularly given policy changes announced since November. We explore the economic implications of alternative scenarios for net migration (see Box 2.3). Box 4.5 below describes how changes in net migration affects our fiscal forecast and explores the implications of the alternative scenarios for the public finances.

4.83 There are also policy-related risks to our fiscal forecast that could have sizeable impacts on the path of fiscal aggregates were they to materialise:

• The fiscal forecast is conditioned on the tax-to-GDP ratio increasing by a further 1.1 percentage points from its current level to close to a post-war high of 37.1 per cent of GDP by 2028-29. Around two-thirds of that increase comes from freezes to personal tax allowances announced since March 2021 which delivers around £7 billion of extra revenue for each year of the freeze. Another seventh comes from the seldom-implemented indexation of fuel duty which delivers around £4.8 billion in additional revenue by 2028-29. If the tax-to-GDP ratio were instead to remain at its 2023-24 level, tax revenues would be £34.3 billion lower in 2028-29.
• There are no detailed departmental spending plans beyond the current Spending Review period which ends in 2024-25. Departmental spending over the subsequent four years of our forecast is assumed to follow two overall envelopes set by the Government. These now imply no real growth in departmental spending per person over the next five years. As shown in Box 4.2, meeting existing commitments on health, defence, schools, childcare, and overseas aid spending would imply a real cut in all other departments budgets of 2.3 per cent per year from 2025-26. Recent history suggests that, as the time comes to allocate the overall spending envelope between departments in a Spending Review, governments tend to top up those envelopes. In the run-up to the two Spending Reviews in November 2015 and October 2021, governments topped up the RDEL envelope by an average of £39 billion and £32 billion per year respectively.

Box 4.5: The impact of migration on the fiscal forecast

Over our five-year forecast, the impact of additional migration on our fiscal forecast is a function of four components:

• the specific fees and charges migrants pay to gain entry to the country. Most migrants, with the exception of asylum seekers and those applying under the EU settlement scheme, pay both the application fees associated with their visa category and the immigration health surcharge (IHS). An average migrant will pay around £1,900 in visa fees and £2,600 on the IHS, and an additional £800 per migrant is paid on the immigration skills charge by sponsoring employers. Total revenues from these sources amount to £4.1 billion per year in our current forecast;
• the general taxes migrants pay as workers, consumers, and residents once they enter the country. As described in Box 2.3, due to new migrants being concentrated among those of prime working age, we estimate they have a slightly higher participation rate than the resident population. Beyond this, we assume new migrants have the same employment, consumption, and residential patterns as residents, and as such pay similar levels of wider taxation, so that their per capita contribution is close to the average UK adult at around £19,500 per year over our forecast;
• the welfare benefits for which migrants are eligible, which are limited for most migrants during at least their first five years in the country. Apart from returning UK or Irish citizens or those who come via humanitarian routes, most new migrants are initially ineligible for most benefits. Eligibility for state pensions requires at least 10 years of qualifying national insurance contributions, meaning almost all new migrants will also initially be ineligible. The impact of new migrants on welfare spending over our five-year forecast is therefore very small;^a and
• the public services they use in the UK. While migrants will consume public services such as education, healthcare, and transport, there is no direct link between the size of the population and the money allocated for departmental spending on public services. Currently, departmental expenditure limits (DELs) for each major public service are fixed by the Government in nominal terms up to the end of the Spending Review period in 2024-25. Thereafter, the Government provides us with an assumption for growth in total current and capital spending, but no detailed department-by-department plans. As discussed in paragraph 4.54, as the Government has not adjusted these limits or assumptions for the increases in population, implied real public service spending per person has fallen since the plans were set out in October 2021.

Therefore, the net fiscal impact of the 350,000 (around 300,000 adults) increase in net migration across our central forecast above Autumn is to:

• increase specific fees and charges paid by those additional migrants by £0.3 billion in 2028-29;
• increase general tax revenue paid by those additional migrants by £6.2 billion in 2028-29;
• leave welfare spending largely unchanged as very few of the new migrants will be eligible by 2028-29;
• leave public services spending largely unchanged, but reduce DEL spending per person by £40 in 2028-29 and increase the pressure on those services; and
• therefore deliver a net reduction in borrowing of around £7.4 billion by 2028-29, taking account of lower spending on debt interest as well.

Fiscal implications of alternative migration scenarios

To explore the fiscal implications of the uncertainty around future migration levels we draw on the higher and lower migration scenarios, described in Box 2.3, in which net migration is 200,000 a year higher or lower than in our central forecast. These scenarios assume that extra migrants have the same levels of participation and hourly productivity as assumed in our central forecast. We then make different assumptions about how spending on public services reacts to levels of net migration.

No adjustment to departmental spending on public services (DEL): In the higher migration scenario, receipts are £18.0 billion higher by the forecast horizon with contributions from both migration specific fees (£0.5 billion) and general taxation (£17.5 billion). With an assumption that migrants receive no welfare and that DEL spending is not adjusted to reflect the increase in population, the only spending impact is £1.9 billion lower debt interest spending from the smaller stock of debt. Overall, borrowing is estimated to be £19.9 billion lower and debt 3.1 per cent of GDP lower in 2028-29. The lower migration scenario is symmetric to this, with lower receipts and spending, meaning that borrowing is £19.9 billion higher and debt 3.1 per cent of GDP higher in 2028-29.

Adjustment to departmental spending on public services (DEL): In the higher migration scenario, if the Government responded to additional migration by keeping departmental spending per person unchanged compared to our central forecast this would require an estimated extra £8.1 billion of expenditure by 2028-29. This may overstate the spending pressures because new migrants tend to be concentrated among those of working age and include fewer children and old people than the resident population. This would reduce the per capita pressures placed on services such as schools and hospitals. Adjusting for this we estimate that providing equivalent service provision to the additional migrants in the high scenario would require an additional £6.1 billion in 2028-29. On this basis, in this scenario a portion of the higher receipts generated by the additional migrants is offset by higher spending, so that while migration still improves the public finances it does so by less than in the unadjusted DEL scenario, with borrowing £13.1 billion lower and debt 2.5 per cent of GDP lower by 2028-29. The lower migration scenario has the opposite effect with borrowing and debt £13.1 billion and 2.5 per cent of GDP higher.

Chart H: Borrowing and PSND ex BoE in the migration scenarios

These scenarios are a simplification and are highly uncertain. They are sensitive to assumptions around the composition of migrants including their age, skill, and average earnings. And they are dependent on the levels of departmental spending the government sets as a response to the higher population. They also only consider the economic and fiscal impacts over our five-year forecast horizon. The overall long-term impact of migration on the public finances is more uncertain. The fiscal impacts of migration are likely to become less beneficial over time, reflecting that after a minimum of 5-years, migrants can apply for indefinite leave to remain and therefore become eligible for welfare benefits. If migrants stay in the UK into older age, there would also be greater pressures on pensions and health spending and lower tax revenues as they retire.

a) Over time some new migrants will convert their residency status to ‘indefinite leave to remain’ and therefore gain the right to access welfare. As the minimum residency required to move to indefinite leave to remain is currently at least five years this falls outside our forecast period. However, our forecasts will capture the cost of any immigrants from previous cohorts who now claim welfare because their claims will be included in the outturn data that provides the starting point for our forecast.

Chapter 5: Performance against the Government’s fiscal targets

Introduction

5.1 This chapter:

• sets out the Government’s fiscal targets and assesses their likelihood of being met on current policy under our central forecast (from paragraph 5.2); and
• considers uncertainty around our fiscal forecast and the risks to the Government meeting its fiscal targets based on historical patterns of shocks of different types, variations in key macroeconomic and fiscal determinants, and alternative scenarios for key forecast judgements (from paragraph 5.16).

The fiscal targets

5.2 The Charter for Budget Responsibility requires the OBR to judge whether the Government has a greater than 50 per cent chance of meeting its fiscal targets under current policy. The first Charter was set in 2011 and it has been updated six times as governments have revised their fiscal targets. The latest version was approved by Parliament on 6 February 2023 and sets out three fiscal targets:

• A ‘fiscal mandate’ that requires public sector net debt excluding the Bank of England as a percentage of GDP to be falling by the fifth year of the rolling forecast period, which is currently 2028-29.
• A supplementary target that requires public sector net borrowing not to exceed 3 per cent of GDP, also by the fifth year of the rolling forecast period in 2028-29.
• An expenditure cap set by the Treasury that requires welfare spending (excluding the state pension and payments most closely linked to the economic cycle) to be contained within a predetermined cap and margin in 2024-25.

5.3 The Charter also identifies a broader set of indicators that the Treasury will consider in its management of fiscal policy, but for which no explicit targets are set. These include:

• Wider public sector balance sheet metrics beyond the narrower debt measure that is targeted in the fiscal mandate, including public sector net financial liabilities and overall public sector net worth.
• Cost of debt metrics that look at the cost of servicing the public debt and its sensitivity to changes in the economic outlook.

The implications of our central forecast

5.4 In our central forecast, the fiscal mandate and the Government’s supplementary target are met by historically small and historically large margins, respectively, and the welfare cap is on course to be missed:

• The fiscal mandate for public sector net debt (excluding the Bank of England) is met by a margin of 0.3 per cent of GDP (£8.9 billion) in 2028-29. This is £4.0 billion lower than the November 2023 forecast of £13.0 billion headroom. The probability of the target being met is assessed as 54 per cent, 2 percentage points lower than our November 2023 forecast.^[43]
• The 3 per cent of GDP borrowing target is met by a margin of 1.8 per cent of GDP (£56.8 billion in 2028-29). This is £4.8 billion lower than our November forecast of £61.5 billion headroom. The probability of the target being met is assessed as 72 per cent, 6 percentage points lower than our November 2023 forecast.
• The welfare cap in 2024-25 is on course to be exceeded by £7.4 billion, a £1.1 billion improvement relative to the £8.6 billion by which it was on course to be missed in November.

5.5 This Budget once again sees the rate of fuel duty temporarily frozen for another year, including the temporary 5p cut introduced at the March 2022 Spring Statement. Stated Government policy thereafter is to reverse the 5p cut and index the duty rate to RPI. Given the Treasury Select Committee’s concerns about our forecasts assuming the rarely implemented RPI indexation of fuel duty rates,^[44] we have also calculated margins against the debt falling and borrowing fiscal targets if fuel duty was not raised in cash terms from its current rate. On this basis, revenue would be £4.8 billion lower than our central forecast in 2028-29, so the fiscal mandate to get debt falling would be met by £4.5 billion (0.1 per cent of GDP), while the borrowing target would be met by £52.0 billion (1.6 per cent of GDP).^[45]

5.6 Fuel duty indexation represents one of a number of risks to our central forecast. Other key risks include the assumptions included in our economy forecast for energy prices (Box 2.2), productivity and migration (Box 2.3), the potential asymmetry of fiscal policy responses to underlying forecast changes (Box 3.1), departmental spending plans beyond the current spending review period (Box 4.2), and uncertainty around the future path of interest rates (Box 4.3).

Table 5.1: Performance against the Government’s fiscal targets

Change in headroom against fiscal targets

Debt falling and borrowing targets

5.7 Our central forecast for pre-measures fiscal headroom is little changed for both the debt falling and borrowing targets relative to our November 2023 forecast. Policy measures announced in the Budget then reduce the headroom modestly. As shown in Chart 5.1:

• Underlying pre-measures forecast changes reduce headroom by £0.8 billion for the debt falling target but increase headroom by £0.3 billion for the borrowing target in 2028-29. For the fiscal mandate, the benefit of lower borrowing in the target year is more than offset by higher financial transactions and the lower starting level of debt.^[46]
• Budget policy measures reduce the headroom against both targets, by £3.2 billion for debt falling and by £5.0 billion for the borrowing target.
• This leaves overall headroom in our central forecast at £8.9 billion (0.3 per cent of GDP) against the debt falling target and £56.8 billion (1.8 per cent of GDP) against the target for borrowing to not exceed 3 per cent of GDP in 2028-29.

Chart 5.1: Fiscal target headrooms changes since November

Changes in the debt-stabilising primary balance over time

5.8 The aim of the Government’s legislated fiscal mandate is for underlying debt (excluding the Bank of England) as a share of GDP to fall by the fifth and final year of the forecast, currently 2028-29. Based on prevailing market conditions and our forecast for financial transactions, stabilising the debt-to-GDP ratio requires the Government to turn what is currently a primary deficit (government borrowing excluding net interest payments) of 1.2 per cent of GDP this year, into a primary surplus of 1.3 per cent of GDP by 2028-29. In this forecast, the Government achieves a primary surplus of 1.6 per cent of GDP in 2028-29, resulting in the fall of 0.3 in the underlying debt-to-GDP ratio. As described in Chapter 4, this is achieved using a mix of reductions in spending and rises in taxes as a share of the economy across the forecast to bring the primary surplus above its debt-stabilising level.

5.9 Chart 5.2 shows the level of the debt-stabilising primary balance since 2010 and in our central forecast. It shows that, over the forecast period, the debt-stabilising primary surplus is significantly higher than it has been at almost any point since 2010 (excluding the pandemic affected year of 2019-20). The primary surplus of 1.3 per cent of GDP required in 2028-29 is 3.4 percentage points higher than the average primary deficit of -2.1 per cent which would have stabilised debt between 2010-11 and 2018-19. This reflects the less favourable growth-corrected interest rate environment and higher levels of financial transactions, as explored in Box 5.1. Given that the forecast debt-stabilising primary balance is a function of future paths of interest rates, economic growth and debt, it is highly uncertain and subject to change most notably from changes in interest rate expectations, as outlined in Box 4.3.

Chart 5.2: Debt-stabilising primary balance since 2010

Box 5.1: The evolution of public sector net debt (excluding the Bank of England) since 2000

As shown in Chart A, since the beginning of the 21^st century, public sector net debt excluding Bank of England (which the Government targets in its fiscal mandate) has increased sharply from 28.4 per cent of GDP in 2000-01 to 84.9 per cent of GDP in 2022-23. The evolution of debt as a share of GDP reflects three factors:

• the size of the primary budget balance – the difference between government spending excluding debt interest payments, and government receipts excluding interest receipts;
• the ‘growth-corrected’ interest rate (‘r-g’) – the difference between the interest rate paid on government debt (which raises the debt-to-GDP ratio) and the growth rate of output (which reduces it); and
• any financial transactions, valuation and classification effects – changes that are not reflected in borrowing but affect the government’s net cash requirement or the value of any liabilities or liquid assets that are captured in net debt.

Chart A: Public sector net debt (ex Bank of England) since 2000

Chart B shows changes in net debt (excluding the Bank of England) divided into these three components.^a It shows that:

• Between 2000 and 2007, the debt-to-GDP ratio rose by 7.4 per cent of GDP, due to a series of primary deficits which averaged 1.3 per cent of GDP after 2001-02. The contributions of financial transactions, and the growth-corrected interest rate, were both small and offsetting over this period.
• Between 2008 and 2010,the financial crisisled to debt rising by close to 35 per cent of GDP. This reflected a sharp rise in the primary deficit, with the cash value of spending rising relative to GDP (reflecting the sharp contraction in nominal GDP), while tax revenues fell even more sharply than GDP. The Government’s policy responses to the crisis in the form of purchases of bank shares, the acquisition of failed banks, and the provision of depositor compensation, meant financial transactions and valuation effects also pushed debt up sharply.
• Between 2011 and 2019,the debt-to-GDP ratio initially rose before levelling off and then falling in the final years of this period. The Government steadily reduced the primary deficit over this period, with the aim of reducing debt, and for most of this period this was supported by a favourable growth-corrected interest rate. Financial transaction and valuation effects (including from the quantitative easing or ‘QE’ operations of the Bank of England and the disposal of financial assets acquired through the financial crisis interventions) reduced debt by 1.7 per cent of GDP per year on average.
• From the end of 2019 to 2023, the combination of the pandemic and energy price shock led to debt rising by 12.6 per cent of GDP. The primary deficit peaked at 13.2 per cent of GDP in 2020-21, mainly due to the sizeable Covid policy support packages. The very sharp fall in growth at the start of the pandemic meant that the growth-corrected interest rate pushed debt up sharply initially, but this reversed even more dramatically when growth bounced back, and interest rates fell.

Chart B: Contributions to changes in the debt-to-GDP ratio (ex Bank of England)

In our central forecast, we expect the debt-to-GDP ratio (excluding the Bank of England) to rise by 3.9 per cent of GDP this year and then by a further 4.4 per cent of GDP between 2023-24 and 2027-28. In the final year, it falls modestly by 0.3 per cent of GDP. This reflects the net effect of:

• financial transactions and valuation effects which push up debt by an average of 1.9 per cent of GDP over the forecast period. These are partly driven by interest-related factors, through losses on the sales and redemption of gilts acquired during QE and held by the Bank of England’s Asset Purchase Facility. The rising stock of net outlays from student loans also adds to debt.^b
• the growth-corrected interest rate which pushes down on debt over the forecast period by an average of -0.4 per cent of GDP per year. The growth-corrected interest rate offers less of a fiscal benefit than during the pre-pandemic period between 2011 and 2019 when it averaged -0.7 per cent of GDP; and
• the primary balance which moves from a primary deficit of 1.2 per cent of GDP in 2023-24 to a surplus of 1.6 per cent of GDP in 2028-29. If realised, this would be the largest primary surplus since the start of this century.

a) Throughout this box, we use March-centred GDP for expressing the primary deficit and financial transactions and valuation effects as a share of GDP to provide a consistent decomposition of debt.

b) In addition to rising net outlays, the change in the undergraduate student loan payment plan from “Plan 2” and “Plan 5” means that a higher share of net outlays are captured within financial transactions and valuation effects rather than the primary deficit.

Headroom against successive fiscal mandates

5.10 The £8.9 billion of headroom against the fiscal mandate in this forecast is the second lowest headroom (excluding the pandemic period) that any Chancellor has had against his fiscal targets since 2010, after the £7.1 billion in today’s terms (£6.5 billion at the time) that Chancellor Hunt had in his March 2023 Budget. Since the OBR was established, governments have had six different fiscal mandates. Chart 5.3 shows headroom against those mandates in successive forecasts. In most cases, Chancellors have held headroom of between £15 billion and £30 billion with an average of £26.1 billion (expressed in terms of today’s GDP). Chancellors have also shown a tendency to spend improvements in headroom while not offsetting this to the same extent when the underlying forecast deteriorates, as explored in Box 3.1.

Chart 5.3: Successive forecasts for headroom against fiscal targets

Welfare cap

5.11 The welfare cap sets a limit on the amount that government can spend on certain social security benefits and tax credits in the final year of a given Parliament, currently 2024-25. It was first introduced in 2014 and has been revised many times since, both substantively and to reflect fiscally neutral reclassifications of spending.

5.12 The welfare cap and margin (a 2.0 per cent allowance above the cap to provide for pressures on, and fluctuations in, capped welfare spending) is on course to be missed by £7.4 billion in our latest forecast (Table 5.2). Spending subject to the welfare cap has been revised down by £1.1 billion relative to our November forecast.

5.13 The Charter stipulates that we must only make a “formal assessment” of performance against the welfare cap in the first Budget of a Parliament, and that instead we should “monitor” progress between those assessments. Only a breach of the cap at the point of a formal assessment triggers Parliamentary processes.^[47]

Table 5.2: The welfare cap and margin

Broader fiscal indicators

5.14 The Charter commits the Treasury to monitoring a broader set of indicators that provide a more comprehensive picture of the public sector balance sheet and consider the affordability of the stock of public debt. This is “with the aim of supporting the achievement of the fiscal objectives”. In addition to public sector net debt (PSND), both including and excluding the Bank of England (ex BoE), the broader balance sheet metrics we forecast include public sector net financial liabilities (PSNFL) and public sector net worth (PSNW). To assist in the assessment of the affordability of debt, we produce forecasts of debt interest costs both as a share of GDP and of government revenue.

5.15 The broader fiscal indicators identified in the Charter are not formal targets, so we do not assess performance against them. Chart 4.18 in the previous chapter showed that all balance sheet metrics are marginally improving by the end of the forecast period, with the wider balance sheet measures of net liabilities peaking earlier and falling more quickly compared with PSND ex BoE. To facilitate further monitoring, Table 5.3 presents those that feature in our forecast in a dashboard that shows: first, their levels and how these compare with the median that prevailed from 1967-68 to 2006-07 (the four decades preceding the financial crisis, before debt ratcheted higher as a result of it); and second, whether they are improving or deteriorating in each year of the forecast.

5.16 The dashboard shows that the:

• Balance sheet (stock) measures are all currently in a much worse position across the forecast than the pre-2007 median, which can be seen in the sea of red in the top section of the top panel. These are all on an improving path by the end of the forecast period as shown by the bottom panel. The path of PSND ex BoE over the forecast is explored further in Box 5.1.
• Cost of debt (flow) measures also peak at historically high levels early in the forecast – indicated by red cells – with net interest costs peaking at 3.0 per cent of GDP and 7.5 per cent of revenue in 2023-24. This is mainly due to higher debt interest spending on index-linked gilts as well as the impact of higher Bank Rate. This is explored further in Box 4.3. Thereafter, net interest costs decrease at a historically fast pace as RPI inflation falls. As a share of GDP, net interest costs fall to below their pre-2007 level both as a share of GDP and as a share of receipts.

Table 5.3: Dashboard of balance sheet and fiscal affordability indicators

Recognising uncertainty

5.17 The headroom that the Government retains against its fiscal rules on our central forecast, and its overall fiscal stance in the medium term, need to be seen in the light of the potential risks to the fiscal outlook. The succession of major shocks that the UK and the global economy have faced since the start of the century has underscored the importance of understanding the uncertainties around a central forecast. The OBR is required to assess whether the Government has a better-than-even chance of meeting its fiscal targets, which we do by producing a median forecast relative to which the outturn is equally likely to be higher or lower than predicted based on currently stated policies.

5.18 We use several analytical tools to illustrate the risks around our central forecast, including:

• fan charts that reflect the chances of shocks of different sizes (through stochastic simulations drawing on historical experience^[48]) to illustrate the uncertainty around our assessment of the probability of the Government meeting its fiscal targets;
• sensitivity analysis that illustrates the vulnerability of the Government’s debt and borrowing targets to changes in key forecast outcomes including growth, inflation, and interest rates; and
• alternative scenarios, which consider the economic and fiscal implications of an adverse shock to one or more of our central forecast assumptions, in this case concerning the paths for energy prices and migration.

Fan charts

5.19 Our fan charts are based on stochastic simulations and allow us to assess the probability of the Government meeting its fiscal targets. Chart 5.4 shows the probability distribution around our forecast of PSND (excluding the Bank of England) and PSNB. It shows that, based on currently stated government policy, there is a 54 per cent chance of underlying debt falling as a share of GDP in 2028-29 and 72 per cent chance of PSNB being less than 3 per cent of GDP in 2028-29. Relative to our November forecast, the chance of debt falling is slightly lower (down from 56 per cent) as is the chance of meeting the borrowing target (down from 78 per cent).

5.20 In this forecast, we have adjusted our methodology for calculating the probability of hitting the fiscal targets to place a greater emphasis on shocks from the last 25 years. This is informed by our view that the number and severity of shocks in the forecast is more likely to resemble recent history rather than the total sample period used for our stochastic simulations, which stretches back to 1955. As recent years have seen a greater variation in borrowing and debt, this leads to a wider variation in our fan charts of possible outcomes. This has made no difference to our estimate on meeting the mandate.

Chart 5.4: Fan charts for PSND (excluding Bank of England) and PSNB

Sensitivities

5.21 Our sensitivity analysis estimates what would need to happen to key forecast parameters and judgements to reduce the headroom against different targets to zero (a ‘test to failure’ or ‘reverse stress test’). In the context of the current rules, we consider the sensitivity of the change in the debt-to-GDP ratio to changes in the economy-wide effective tax rate, the effective interest rate on government debt, nominal GDP growth and inflation. Given the Government meets the supplementary target to reduce borrowing below 3 per cent of GDP by a relatively large margin (1.8 per cent of GDP), we have omitted any sensitivity analysis for this target.

The change in the debt-to-GDP ratio

5.22 We use our fiscal ready-reckoners to calibrate several possible adverse surprises, relative to our central forecast, that would be sufficient to negate the 0.3 per cent of GDP year-on-year fall in debt (excluding the Bank of England) in 2028-29.^[49] It could fall to zero if:

• Nominal GDP growth was 0.3 percentage points lower in 2028-29, shown by the red diamond in the upper-left panel of Chart 5.5. This is equal to the amount by which GDP growth undershot our November forecast in 2023.
• The effective tax rate was 0.3 percentage points lower in 2028-29, shown by the red diamond in the upper-right panel of Chart 5.5. This is equivalent to the cost of the 2 percentage point cut in national insurance contributions in this forecast.
• Effective interest rates on central government debt were 0.3 percentage points higher at 3.4 per cent in 2028-29, shown by the red diamond in the lower-left panel of Chart 5.5. This change in rates is broadly equivalent to the volatility in gilt rates since we closed our forecast for market determinants.
• RPI inflation was 1.3 percentage points higher in 2028-29, shown by the red diamond in the lower-right panel of Chart 5.5. This would be an upward shock to RPI of one-tenth of the size of the inflation shock we experienced in 2022-23.

Chart 5.5: Sensitivities

Scenarios

5.23 In Chapter 2 we outline an array of global and domestic risks that could push the public finances off course. These include risks related to our assumptions for productivity, inflation, migration, and energy prices. In this EFO, we have assessed the fiscal implications of two of these: migration and energy prices.

• Box 4.5 outlines the impacts of higher or lower migration on the public finances. It shows that over a five-year horizon, migration is probably beneficial for the public finances, but the extent of the benefit is sensitive to assumptions around the age and skill profile of migrants, and the spending response from Government.
• In this section, we consider the fiscal implications of a sharp increase in energy prices and a supply-chain shock, drawing on the economic assumptions outlined in Box 2.2 around the potential for further instability in the Middle East.

Energy price shock scenario

5.24 In Box 2.2, we explored the economic implications of a wider disruption to trade flows from and through the Middle East than assumed in our central forecast. In this scenario, oil, gas and electricity prices rise to 75 per cent above our central forecast for a year. Disruptions to supply chains for goods also return to their pandemic peak before falling back. As a result, CPI inflation peaks at 7.4 per cent in the beginning of 2025, before dropping below target in 2026. Bank Rate rises to around 7 per cent then falls and remains below our central forecast for the remainder of the scenario. Nominal GDP rises in the near-term due to a higher deflator but falls back close to our central forecast in 2025-26 as real GDP falls. By the forecast horizon, real GDP recovers but a permanently higher deflator means the level of nominal GDP is 2.2 per cent higher by 2028-29.

5.25 As shown by the solid yellow lines in Chart 5.6, the shock to the economy leads to a significant deterioration in the public finances. Over the 5-year forecast period borrowing is £23.1 billion higher per year on average though the gap has closed by the end of the period. Debt peaks 2.9 per cent of GDP higher than in the central forecast before falling to 0.8 per cent higher as a share of GDP by the end of the period. This is driven by:

• Higher welfare and pensions spending (£21.6 billion higher by the forecast horizon), reflecting higher CPI which is used for indexation.
• Higher debt interest spending (£13.3 billion higher by the forecast horizon), due to both higher interest rates and RPI inflation which raise the cost of servicing the Government’s existing debt. This is further compounded by higher government borrowing which raises debt interest further.
• This is offset partly by higher tax receipts. Offshore corporation tax and electricity generator receipts are boosted temporarily by higher oil, gas, and electricity prices. Over the medium-term, the permanently higher deflator boosts nominal tax bases and increases the effective tax rate (via ‘fiscal drag’) and so raises receipts. By the forecast horizon, receipts are £31.9 billion higher. ^[50]

5.26 These results do not make any assumptions around the profile of departmental spending (DEL). If departmental budgets were adjusted to preserve their real spending power then this would lead to a further £20.4 billion increase in borrowing by the forecast horizon as shown in the dotted yellow line. Debt would peak at 4.6 per cent of GDP higher than in the central forecast and fall to 3.5 per cent higher by the end of the period.

5.27 In each case despite the higher paths for borrowing and debt the fiscal mandate would continue to be met as debt is falling as a share of GDP at the end of the period. This is due to the assumption in this scenario that the energy price shock is temporary and subsequently there is faster nominal GDP growth at the forecast horizon as real GDP recovers from the shock.

Chart 5.6: Borrowing and debt (ex BoE) in the energy price shock scenario

Annex A: Detailed tables

A.1 This annex contains summary tables providing a detailed breakdown of the economy and fiscal forecasts described in this Economic and fiscal outlook. We also include changes since our November 2023 Economic and fiscal outlook. These tables include:

• a detailed summary of our economy forecast and key determinants of the fiscal forecast;
• public sector current receipts and individual taxes;
• contributions to total managed expenditure;
• the main fiscal aggregates; and
• sources of year-on-year changes in public sector net debt.

Table A.1: Economy forecast

Table A.2: Economy forecast: changes since November

Table A.3: Determinants of the fiscal forecast

Table A.4: Determinants of the fiscal forecast: changes since November

Table A.5: Current receipts