Foreword

This Economic and fiscal outlook (EFO) sets out our central forecast for the five years to 2028-29, taking account of recent data and government policies announced up to and including the Autumn Statement 2023. 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 hard work and expertise. We have also drawn on the work and expertise 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 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 for their engagement, expertise, and insights.

Outside government we have held useful discussions with the Bank of England, the Society of Motor Manufacturers and Traders, the Greater London Authority, the National Institute of Economic and Social Research, the Institute for Fiscal Studies, the Resolution Foundation and the Institute for Government. Our assessment of a number of policies that were outlined or announced in advance of the Autumn Statement has greatly benefitted from discussions with: Tony Wilson, Kate Barker, Toby Nangle, Barry Kenneth, Maria Nazarova-Doyle, Madeleine Sumption, Alan Manning and Tessa Hall. We are also grateful to our advisory panel for sharing their insights and useful discussions and to Hamish Low for his helpful advice on labour supply elasticities.

The date for the Autumn Statement and this forecast was announced on 5 September, well in advance of the required ten weeks’ notice. This met the agreed process for fiscal events, as outlined 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) which was recently updated and published on our website on 9 November.

We published the timetable of the key stages of the forecast on 21 September, once it had been agreed by signatories of the MoU. Overall, the forecast process for this EFO proceeded very smoothly, marking a welcome return to previous practice after a series of fiscal events with disrupted timetables. The timetable was very largely adhered to at each stage and proceeded as follows:

• OBR staff prepared an initial economy forecast, drawing on data released since our previous forecast in March 2023 and incorporating our preliminary judgements on the outlook for the economy. This first economy forecast was sent to the Chancellor on 21 September.
• Using the economic determinants from this forecast (such as the components of nominal income and spending, unemployment, inflation, and interest rates), we commissioned updated 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 4 October.
• As the process continued, we identified further key judgements that we would need to generate our full economy forecast. Where we thought it would be helpful, we commissioned analysis from the relevant teams in the Treasury to inform our views and discussed forecast issues, though not specific judgements, with experts from external organisations. The BRC then agreed further judgements, allowing the production by OBR staff of a second economy forecast, which was sent to the Chancellor on 10 October.
• 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 follow up our requests for further analysis, methodological changes, and alternative judgements from the previous round. We sent our second fiscal forecast to the Chancellor on 19 October.
• In parallel, as the Autumn Statement involved a set of policy measures that we deemed could have an impact on our forecast for potential output, we undertook a process of engagement and analysis 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 March 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.
• We then produced a third and final pre-measures economy forecast, in which we took on the latest data and incorporated judgements embodied in our fiscal forecast. This economy forecast included energy and financial market data based on the average over the 10 working days to 11 October and was sent to the Treasury on 25 October. The associated fiscal forecast was sent to the Chancellor on 31 October.
• 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 2 November, although some of the supporting analysis provided by the Treasury was received beyond the agreed deadline. We incorporated this package into a preliminary post-measures forecast, in order to provide an early view on the effect of Autumn Statement measures on the economy and public finances, which we sent to the Chancellor on 7 November. This forecast round was produced using our internal ready-reckoner models (rather than being sent to departmental forecasters). We recently published an updated ready reckoner tool on our website that closely aligns to the actual tools we use for this part of the process.
• In line with the agreed timetable, on 9 November the Treasury provided the final package of measures that would cause movements in our economy forecast. We sent the resulting final economy forecast to the Treasury on 14 November and a near-final fiscal forecast on 15 November. Final policy decisions were provided by the Treasury on 16 November and our forecast was then finalised on 17 November and sent to the Treasury on the same day. We were notified of a change to one policy decision on 18 November, which would have had a negligible impact on our forecast had we been able to include it.
• 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 17 November. This allowed the Treasury to prepare the Chancellor’s statement. We also provided pre-release access to the full and final EFO on 20 November.

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. Richard Hughes met with the Chancellor on 6 and 15 September to discuss early forecast judgements. The BRC also met with the Chancellor twice, to discuss the forecast over the course of its production (on 10 October and 2 November). 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 17 November.

Our non-executive members, Bronwyn Curtis OBE, Dame Susan Rice and Baroness Hogg 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

1.1 The economy has proved to be more resilient to the shocks of the pandemic and energy crisis than anticipated. By the middle of this year, the level of real GDP stood nearly 2 per cent above its pre-pandemic level and around 3 per cent above our March forecast. But we now expect the economy to grow more slowly over the forecast period, leaving the level of real GDP only ½ a per cent higher in the medium term than in our March forecast. Inflation is expected to be more persistent and domestically fuelled than we previously thought, falling below 5 per cent by the end of this year but not returning to its 2 per cent target until the first half of 2025, more than a year later than in March. Markets now expect interest rates will need to remain higher for longer to bring inflation under control. Despite the more challenging outlook for the real economy, higher inflation leaves nominal GDP nearly 5½ per cent higher by the start of 2028 than we forecast in March.

1.2 More persistent, domestically driven inflation boosts nominal tax revenues compared to March. But it also raises the cost of welfare benefits, and higher interest rates raise the cost of servicing the Government’s debts. It is mainly due to the Chancellor’s decision to leave departmental spending broadly unchanged that higher inflation and other forecast changes reduce borrowing by £27 billion in 2027-28 compared to our March forecast. The Chancellor spends this windfall on cuts in National Insurance Contributions, permanent up-front tax write-offs for business investment, and a package of welfare reforms, which together provide a modest boost to output of 0.3 per cent in 5 years. He still meets his target to get debt falling as a share of GDP in 5 years’ time by an enhanced margin of £13 billion, but mainly thanks to the rolling nature of the rule giving him an extra year to get there. And while personal and business tax cuts reduce the tax burden by ½ a percentage point, it still rises in each of the next 5 years to a post-war high of 38 per cent of GDP.

Economic outlook

1.3 The economy recovered more fully from the pandemic and weathered the energy price shock better than anticipated. ONS revisions now show that the economy recovered its pre-pandemic level at the end of 2021 and was 1.8 per cent above it in mid-2023, rather than 1.1 per cent below as we had assumed in March. Revisions to growth rates in the last couple of years were more muted, but the economy has so far also proven more resilient than we expected in the face of higher energy prices, inflation, and interest rates, with cumulative growth nearly 1 percentage point stronger in the first half of 2023 than our March forecast. The combined effect of the historical revisions and latest outturns leaves the level of real GDP at the start of this forecast almost 3 per cent higher than we thought in March. But survey data suggest that much of that unexpected economic strength can be attributed to a modest degree of excess demand, rather than the excess supply we anticipated in March.

1.4 We therefore expect inflation to remain higher for longer, taking until the second quarter of 2025 to return to the 2 per cent target, more than a year later than forecast in March. Inflation is also more domestically fuelled with a more positive starting output gap and stronger nominal wage growth more than offsetting the faster-than-expected decline in gas prices. From a peak of 10.7 per cent in the last quarter of last year, CPI inflation is now expected to fall to 4.8 per cent in the final quarter of 2023. As a moderate degree of spare capacity in the economy opens up and gas prices fall further, inflation dips slightly below the 2 per cent target between 2025 and 2027, before returning to it at the forecast horizon.

Chart 1.1: CPI inflation

1.5 More persistent inflation means markets expect interest rates to be more than a full percentage point higher than we assumed in March. Bank Rate reached a 15-year high of 5.25 per cent in August 2023, around 100 basis points above our previous forecast. And markets now expect Bank Rate to settle at 4 per cent by the end of the forecast, rather than fall to 3 per cent as we assumed in March. Gilt yields are between 100 and 150 basis points higher across the yield curve, with 10-year gilt yields also at a 15-year high of 4.5 per cent, compared to 3.3 per cent at the time of our last forecast. Past rises in interest rates are still to have their full effect on the economy, with the increased share of fixed rate mortgages in recent years slowing the pass-through of higher mortgage rates to household incomes. The Bank of England estimates that over half the impact from two years of interest rate increases is still to be felt.^[1]

1.6 In our central forecast, we have revised down our estimate of the medium-term potential growth rate of the economy to 1.6 per cent, from 1.8 per cent in March. The revision is largely driven by a weaker forecast for average hours per worker, which we now expect to fall in the medium term, rather than holding flat. This largely reflects our reassessment of the effect of demographic shifts in the composition of the working population toward younger and older age groups who work shorter hours on average. We now also forecast a rising retirement rate for capital, which is more consistent with recent technological trends and dampens our forecast for capital deepening and productivity. On top of these factors, recent data and historical revisions point to a weaker near-term outlook for total factor productivity growth. The impact of these downward revisions on the medium-term level of GDP are only partly offset by higher migration, stronger business investment, lower energy prices, and the policy measures which further boost labour supply and business investment in this forecast.

1.7 Slower growth from a higher starting point means that, in our central forecast, the level of real GDP in 2027 is only 0.6 per cent higher than March. Business surveys and consumer confidence measures point to continued weak growth over the second half of 2023. The ONS estimates that real GDP growth was zero in the third quarter, and we expect growth of only 0.1 per cent in the fourth quarter of 2023. On an annual basis, real GDP growth slows from 4.3 per cent in 2022 to 0.6 per cent this year, and 0.7 per cent next year. Squeezed real wages, higher interest rates, and unwinding government support all weigh on economic activity, opening up a moderate degree of spare capacity over the next three years. Growth then picks up to 1.4 per cent in 2025 and an average of 1.9 per cent between 2026 and 2028 as the squeeze on real wages eases and interest rates fall back, closing the output gap. Cumulative real growth from 2023 to 2027 is 2.4 percentage points lower than forecast in March due to a weaker forecast for potential output growth and less spare capacity in the economy at the start of the forecast period.

Chart 1.2: Real GDP

1.8 Real GDP per person remains 0.6 per cent below its pre-pandemic peak and in the central forecast only recovers that peak at the start of 2025. Weak economic growth despite higher levels of net migration, that reached 606,000 in 2022, means that real GDP per person is expected to continue to fall in the second half of 2023. It then recovers as GDP growth picks up while net migration falls back towards its assumed long-run level of 245,000 by 2026-27. The decline in migration is partly due to the tighter restrictions on international students bringing dependants and increases in immigration fees announced since March.

1.9 We expect the Chancellor’s policy package in this Autumn Statement to provide a modest boost to demand in the near term and a small but lasting benefit to the supply side of the economy. The near-term fiscal outlay temporarily pushes up aggregate demand relative to supply by 0.1 per cent at the peak of its impact in 2025-26. Three elements of the policy package are also estimated to raise potential output by 0.3 per cent of GDP in 2028-29. The switch from temporary to permanent full expensing removes the incentive to bring forward business investment but raises the overall capital stock in 2028-29 by 0.2 per cent. This increases potential output by 0.1 per cent in 2028-29 and just under 0.2 per cent in the long run. A package of welfare and other measures targeting the inactive and long-term unemployed is estimated to raise employment by 50,000 and hours worked by 28,000 in full-time equivalent (FTE) terms by 2028-29. But the biggest increase in potential output comes from the cut in National Insurance Contribution (NIC) rates which is estimated to raise employment by 28,000 and total hours worked by 94,000 in FTE terms by raising the post-tax gains from work.

1.10 In our central forecast, unemployment rises to 1.6 million people (4.6 per cent of the labour force) in the second quarter of 2025. This peak in unemployment is around 85,000 people (or 0.2 percentage points) higher, and a year later, than we expected in March. The weakening in labour demand reflects rising interest rates and slower GDP growth opening up a degree of spare capacity in the economy. Unemployment then falls back to its assumed structural rate of 4.1 per cent by the forecast horizon, as Bank Rate falls and the spare capacity in the economy is taken up. The employment rate is currently still 1 percentage point below its pre-pandemic peak, but by the end of the forecast period is expected to be 0.3 percentage points higher than we expected in March, due to the Chancellor’s policy measures and our assumption that some of the recent strength in employment is structural.

1.11 Living standards, as measured by real household disposable income (RHDI) per person, are forecast to be 3½ per cent lower in 2024-25 than their pre-pandemic level. While this is half the peak-to-trough fall we expected in March, it still represents the largest reduction in real living standards since ONS records began in the 1950s. RHDI per person recovers its pre-pandemic level in 2027-28, something not achieved in our March forecast, as resilient labour incomes begin to steadily outmatch slowing inflation. We estimate that the reduction in the rate of NICs announced in the Autumn Statement will boost real household incomes by around 0.5 per cent at the end of the forecast.

1.12 Nominal GDP is forecast to be around 5½ per cent higher in the medium term than expected in March. This is driven by a combination of historical revisions and our forecast for higher growth in the GDP deflator which more than offsets our weaker forecast growth in real GDP. Our March outlook assumed that the spike in inflation was mostly an externally driven terms-of-trade shock brought about by a spike in global energy prices, with limited implications for the GDP deflator and nominal GDP growth. But a more dramatic fall back in energy prices and tighter labour market than anticipated in March, suggest the main drivers for inflation have switched from external to domestic pressures, especially nominal earnings (Chart 1.3). As a result, our forecasts for growth in labour incomes, nominal consumption, and profits have all been revised up significantly. Stronger growth in these key tax bases, coupled with tax thresholds that remain frozen over the next five years, drive a significant increase in our tax receipts forecast.

Chart 1.3: Energy prices and nominal earnings

Fiscal outlook

1.13 Borrowing was £19.8 billion lower in the first half of this financial year than we forecast in March, due almost entirely to stronger receipts growth. Central government receipts in the first half of 2023-24 were £14.7 billion (3.3 per cent) above our March forecast while central government spending has been only marginally higher at £0.6 billion (0.1 per cent). The strength in receipts compared to our forecast has been seen across the major taxes – income tax, VAT, and corporation tax – and mainly reflects higher inflation and earnings this year compared to our March forecast. We expect continued receipts strength in the remainder of the year, but also much higher debt interest spending reflecting higher RPI inflation and Bank Rate than we forecast in March. As a result, before the impact of the measures in this Autumn Statement, public sector net borrowing for 2023-24 as a whole is expected to be £15.8 billion (13.6 per cent) lower than in our March forecast.

1.14 The medium-term fiscal outlook has also improved materially compared to March, with pre-measures borrowing forecast to be £26.8 billion lower in 2027-28 (Chart 1.4). More persistent, domestically generated inflation coupled with frozen tax thresholds raises pre-measures receipts by £59.0 billion by 2027-28, with just over half coming from higher income tax and NICs receipts. The £32.2 billion increase in pre-measures spending is more muted as higher inflation and interest rates increase welfare spending and debt interest costs, but not departmental and other spending which remains largely unchanged. The reduction in 2027-28 pre-measures borrowing relative to our March forecast is therefore mainly a reflection of a £19.1 billion erosion in the real value of departmental spending.

1.15 The Chancellor spends almost all of the pre-measures forecast improvement in borrowing between 2023-24 and 2027-28 on his Autumn statement measures, leaving post-measures borrowing largely unchanged. Policy measures announced since our March forecast raise borrowing in every year and by £26.6 billion in 2027-28, which includes:

• a significant reduction in rates of National Insurance Contributions (NICs), including a 2p cut in the main rate of employee NICs, which costs £10.4 billion by 2027-28;
• a permanent 100 per cent capital allowance for qualifying business investment (known as ‘full-expensing’) which costs £9.1 billion in 2027-28;
• a package of reforms to welfare and health services designed to increase labour market participation which save £0.6 billion, wider changes to public spending which cost £8.2 billion, and a set of other tax decisions which raise revenue by £2.0 billion, in 2027-28; and
• the indirect effects of these and other policies on the economy, which boost demand in the near term and labour supply and business investment in the medium term, adding a further £1.5 billion to borrowing in 2027-28 as higher debt interest spending (£3.6 billion) outweighs reductions from other economic effects (£2.1 billion).

Chart 1.4: Public sector net borrowing: change since March

1.16 Tax changes in this Autumn Statement reduce the tax burden by 0.7 per cent of GDP but it still rises in every year to a post-war high of 37.7 per cent of GDP by 2028-29. Income tax increases explain most of the increase in this forecast, rising from 10.2 per cent of GDP this year, to 11.3 per cent in 2028-29 driven by threshold freezes and strong nominal earnings growth. By 2028-29, frozen thresholds result in nearly 4 million additional workers paying income tax, 3 million more moved to the higher rate, and 400,000 more paying the additional rate. VAT and corporation tax also rise from 6.4 and 3.4 per cent of GDP this year to 6.5 and 3.6 per cent of GDP in 2028-29, as consumption shifts back towards standard VAT-rated goods and companies face the full-year effects of the rise in the main corporation tax rate.

1.17 Spending is forecast to fall steadily as a share of the economy from 44.8 to 42.7 per cent of GDP over the forecast but remains 3.1 per cent of GDP above its pre-pandemic level. Debt interest payments fall from the current very elevated levels as RPI inflation falls but remains 1 percentage point higher than the post-war average. Welfare spending rises as a share of GDP until 2025-26 after which it falls slightly as CPI inflation returns to more normal levels. The single largest contribution to the decline in spending comes from departmental spending falling from 19.2 to 18.1 per cent of GDP over the forecast.

1.18 Despite an increase of £4.1 billion a year on average in this Autumn Statement, higher inflation means the real value of departmental spending is £19.1 billion lower by 2027-28 than our March forecast. The current Spending Review period ends in 2024-25, and beyond this departmental spending follows two overall envelopes set by the Government in which day-to-day spending grows by 0.9 per cent a year in real terms and capital spending is fixed in cash terms. As previous spending reviews have approached, governments have topped up annual day-to-day spending envelopes significantly: by £39 billion (14 per cent) on average in the year up to the November 2015 Spending Review, and by £32 billion (8 per cent) in the October 2021 Spending Review. The outlook for departmental spending is therefore a significant and growing risk to our forecast.

1.19 Borrowing is forecast to fall steadily from 5.0 per cent of GDP this year to 1.1 per cent of GDP by 2028-29, which would be its lowest level since 2001-02. This is little changed from our forecast in March as the reduction in the pre-measures forecast is almost entirely offset by the cost of the Autumn Statement measures. Most of the 3.5 per cent of GDP decline in borrowing over the forecast period comes from the increase in income tax and NICs receipts driven by higher earnings and fixed tax thresholds (-1.0 per cent of GDP), the reduction in departmental expenditure as a share of GDP (-1.1 per cent of GDP), and debt interest costs falling back from their peak (-0.5 per cent of GDP). In cash terms, borrowing is forecast to fall from £128.3 billion in 2022-23, to £123.9 billion this year, to £35.0 billion by 2028-29.

Chart 1.5: Public sector net borrowing

1.20 Underlying public sector net debt (excluding the Bank of England) rises from 84.9 per cent of GDP last year to a peak of 93.2 per cent of GDP in 2026-27, then falls slightly in the final two years to 92.8 per cent of GDP in 2028-29. The path of underlying debt is little changed as a share of GDP from our March forecast after taking account of historical revisions to the level of nominal GDP. Headline public sector net debt rises from 97.9 per cent of GDP this year to 98.6 per cent of GDP in 2024-25, then falls to 94.1 per cent of GDP in 2028-29.

Chart 1.6: Public sector net debt excluding the Bank of England

Performance against the Government’s fiscal targets

1.21 The Government’s primary fiscal target is for public sector net debt excluding the Bank of England to fall in the final year of the forecast (2028-29 in this forecast). On our central forecast, this rule is met by a margin of £13.0 billion (0.4 per cent of GDP), up from the £6.5 billion (0.2 per cent of GDP) margin in our March forecast where the target year was 2027-28. This £6.5 billion improvement in headroom reflects:

• Underlying forecast changes which improve the headroom in 2027-28 by £19.6 billion. This is largely explained by the lower borrowing in cash terms in 2027-28, which is partly offset by higher financial transactions (reflecting higher redemption losses on the sale of gilts from the Asset Purchase Facility).
• Rolling on the target year to 2028-29 increases the pre-measures headroom by £4.9 billion as the underlying tax burden continues to rise while spending continues to fall as a share of GDP, further reducing borrowing in this year.
• Policy measures that reduce the headroom by £18.0 billion, offsetting around three-quarters of the pre-measures improvement in headroom.

Chart 1.7: Underlying debt falling headroom: changes since March

1.22 Headroom of £13.0 billion is considerably lower than the average of £29.7 billion that Chancellors have held against their fiscal rules since 2010. Our forecast again incorporates £6.2 billion of extra revenue in 2028-29 from the Government’s stated policy of increasing fuel duty rates in line with RPI inflation and the reversal of the ‘temporary’ 5p cut. If, like all Chancellors since 2011, rates are instead held at the current rate then more than 43 per cent of the headroom in 2028-29 would be removed and debt would no longer be falling in 2027-28. The current headroom is also less than it would have cost to maintain the real value of departmental spending at the same level as in our March forecast.

1.23 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 – £61.5 billion, or 1.9 per cent of GDP – up by £22.3 billion from March. History suggest that these margins would be consistent with a 56 and 78 per cent chance of meeting the mandate and supplementary target respectively.

Risks and uncertainties

1.24 Over recent years large shocks and their aftermath have resulted in significant revisions to our economic and fiscal forecasts from one fiscal event to the next. The same is true in this forecast with the net effect of historically large changes in energy prices, wage growth, and interest rates since March proving to be positive for the public finances. We continue to emphasise the uncertainties around our forecasts with the possibility that any of our key judgements could prove significantly too optimistic or pessimistic.

1.25 Revisions to two economic judgments are particularly important in this forecast, so we present alternative scenarios for each over the next five years:

• Imported versus domestic inflation where we explore a scenario where inflation is higher for longer (remaining above 4 per cent over 2024) and the Bank of England raises rates to a peak that is 1½ percentage points above the level assumed in our central forecast. Were this purely an imported inflation shock, we assess that borrowing would be £23.2 billion higher by 2028-29. However, if the inflation fed fully into nominal GDP and wages, borrowing would be £23.0 billion lower. But that improvement would largely come from eroding the real value of departmental spending. If government looked to maintain their real spending power, two-thirds of the improvement would disappear.
• Productivity growth where annual growth is ½ a percentage point above or below our central estimate of around 1 per cent over the next five years. Productivity growth of 1½ per cent, i.e., a return toward its pre-financial crisis rates, would reduce borrowing by £46.0 billion in 2028-29 while growth falling back to its post-financial crisis average of ½ per cent would increase borrowing by £40.5 billion.

1.26 There are also risks to our forecast associated with the impact or implementation of other stated policies and aspirations. On the revenue side, these risks include: the planned 6p increase in the fuel duty rate in March 2024, a consultation on a new vaping duty, and a consultation on a carbon border adjustment mechanism. On the spending side, risks include: the reduction in the real value of departmental spending, the squeeze on those budgets beyond the current Spending Review (where recent history points to governments’ topping up day-to-day spending envelopes by more than £30 billion a year), and the Government’s ambition to raise defence spending to 2.5 per cent of GDP at an unspecified date (which would cost an additional £16.1 billion if implemented in 2028-29).

Chapter 2: Economic outlook

Introduction

2.1 This chapter describes our latest economy forecast 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 2.10);
• how historical revisions, the starting output gap and our forecasts for labour supply, business investment and total factor productivity determine the path of potential output (from paragraph 2.15);
• prospects for real GDP and the output gap (from paragraph 2.26);
• the outlook for the labour market (from paragraph 2.31);
• our forecasts for the housing market, household incomes, the saving rate, and consumption (from paragraph 2.36);
• the outlook for the trade, current account and sectoral balances and nominal GDP (from paragraph 2.44); and
• how our forecast compares to recent external forecasts (from paragraph 2.49).

Table 2.1: Key economy forecast assumptions and judgements

Key economy forecast assumptions

Commodity prices

2.2 Market expectations for natural gas prices have fallen in both the very near term and the medium term since our March 2023 forecast. Our forecast is now conditioned on wholesale gas prices averaging £1.20 per therm in 2024 and 2025. This is 40 and 20 pence a therm lower respectively than we assumed in March. From 2026, we hold gas prices constant in real terms, which is still around twice the pre-pandemic average of 50 pence per therm. The recent fall in prices (left panel, Chart 2.1) partly reflects a rapid rise in liquified natural gas imports, especially from the US and Qatar, compensating for the loss of Russian imports. The unusually warm winter in 2022-23 also allowed European countries to fill gas reserves above target levels ahead of the coming winter. Despite recent relative stability in the futures market, further supply disruptions or a cold winter remain key risks to gas prices.

2.3 Since our March forecast, expectations of Brent crude oil prices for 2024 have increased 8.3 per cent to $82 a barrel, amid production cuts by OPEC+ members including Russia. In sterling, oil prices for 2024 have risen by 7.9 per cent to £68 a barrel reflecting the slight strengthening in the pound since our March forecast. The price of crude oil has been volatile in recent months, peaking at $97 dollars a barrel on 27 September, 8.9 per cent higher than the price when our current forecast closed to new data (right panel, Chart 2.1). A higher oil price is a key risk given the ongoing conflict in the Middle East. A sustained period of higher oil prices would raise inflation and reduce consumption and productivity.

Chart 2.1: Gas and oil prices

Monetary policy and gilts

2.4 After 13 increases in Bank Rate since the start of 2022, markets now assume the Bank of England’s policy rate will peak only a little above their current level of 5.25 per cent in the final quarter of 2023.^[2] This is more than 1 percentage points above the 4.3 per cent peak in mid-2023 that we expected in our March forecast (left panel, Chart 2.2). This higher and later peak for Bank Rate is driven by more persistent inflation than expected in March. Bank Rate remains more than 1 percentage point higher on average over the November 2023 forecast period than we assumed in our March forecast and 2.5 percentage points higher than our pre-Ukraine invasion forecast in October 2021.

2.5 Yields on government 10-year gilts have risen to a 15-year high of 4.5 per cent, 1.2 percentage points up from our March forecast. By mid-October, yields were up across all maturities, with the largest increases at the shorter end where 5-year gilt yields were up 1.3 percentage points. Yields were above 4 per cent across the curve, with 25-year gilts yielding 5 per cent, 1.2 percentage points higher than in March (right panel, Chart 2.2). Monthly yields on 25-year gilts last exceeded 5 per cent in 2002, when the net stock of government debt was around one third its current size at 28 per cent of GDP.

Chart 2.2: Bank Rate and gilt yields by year of maturity

Exchange rates and the world economy

2.6 The trade weighted sterling effective exchange rate has strengthened by 4.4 per cent since our March 2023 forecast (left panel, Chart 2.3). As of mid-October, the pound had strengthened slightly (0.3 per cent) against the dollar compared to our March forecast. But it is still down 12.9 per cent from its recent high of $1.40 dollars to the pound in the middle of 2021. Since March, the pound has also appreciated 2.4 per cent relative to the euro and 13.9 per cent against the Japanese yen.

2.7 Global economic conditions have deteriorated since our March forecast. Energy and food markets have faced continued disruption due to Russia’s invasion of Ukraine, while monetary conditions have tightened globally to bring down inflation. GDP growth, especially in advanced economies, is expected to slow this year and next. In line with the IMF’s October 2023 World Economic Outlook projections, we no longer anchor our medium-term outlook for world GDP to the pre-pandemic trend. This means we have downgraded cumulative 2023 to 2027 world GDP growth to 12.8 per cent in our central forecast, 2.4 percentage points lower than in our March forecast (right panel, Chart 2.3). Our world trade forecasts are also consequently lower. Conflict in the Middle East poses an ongoing risk to the global outlook, although commodity prices have so far responded little.^[3]

Chart 2.3: Effective exchange rate and World GDP level

Fiscal policy

2.8 The large-scale fiscal support to households, firms, and public services provided during the pandemic and subsequent energy price shock has now mostly been withdrawn as the acute phase of the pandemic has passed and energy prices receded. That withdrawal, combined with net tax rises (largely through freezes to personal tax thresholds) and a reduction in departmental spending as a share of GDP, reduces borrowing by 4 percentage points over the forecast period and puts the underlying debt to GDP ratio on a declining path. The fall in borrowing is largest in the first two years of the forecast and then moderates. As a result, the boost to demand from the fiscal stance wanes over the forecast period. Relative to March, there is a slower initial fall in borrowing, but from a lower starting level as the public finances are set to improve this year by more than we expected.

2.9 The discretionary fiscal policy measures in the Autumn Statement moderately loosen the fiscal stance and provide a small boost (0.1 per cent of GDP) to demand in the near term (described in Box 2.1, but do not significantly affect the overall impact of fiscal policy on demand over the medium-term or have a material impact on the path of inflation.

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’. We also account for the supply side impact of selected policy measures where credible evidence suggests they are likely to have a material, additional, and durable impact on potential output, as discussed in our recent article.^a

Policies announced in this Autumn Statement are expected to add around £17 billion a year on average to public sector net borrowing between 2024-25 and 2025-26, picking up to £23 billion in 2027-28 before falling back to around £17 billion by 2028-29. As shown by the diamonds in Chart A, these policies raise GDP by an average of just under 0.3 per cent between 2024-25 and 2028-29 through a combination of impacts on demand and supply.

Temporary impacts on demand

The Autumn Statement policy package boosts aggregate demand relative to supply (purple bars) by 0.1 per cent at the peak of its impact in 2025-26. This effect is mainly calculated using fiscal multipliers drawn from the empirical literature,^b though also takes into account a small partly offsetting effect (described below) from the changes to the ‘full expensing’ policy. As usual, we assume the demand impact of the policies tapers to zero over the forecast period, as the Bank of England acts to bring aggregate demand in line with potential output.

Impacts on supply

Several items within the Autumn Statement policy package are expected to have lasting supply-side effects, raising the level of real GDP by 0.3 per cent in 2028-29 (Chart A):

• Cuts to NICs (yellow bars) are estimated to raise employment by 28,000 (under 0.1 per cent) in 2028-29. The increase in post-tax income is also likely to boost the hours worked by existing employees. The total increase in hours worked from new and existing employees is estimated to be 0.3 per cent or 94,000 in full-time equivalent terms.^c
• A package of welfare and other measures (green bars) that we estimate will raise employment by 50,000 (over 0.1 per cent) in 2028-29. There is a small positive impact on GDP of only 0.04 per cent as we assume entrants are likely to join on lower-than-average hours and earnings. These measures are described in more detail in Chapter 3.
• The move from temporary to permanent 100 per cent capital allowances – known as ‘full expensing’ – (blue bars) generates a net additional £13.9 billion of cumulative real business investment over the forecast period (further discussed in Box 2.4). In contrast to the previous temporary policy, this raises the capital stock by 0.2 per cent in 2028-29 and potential output by 0.1 per cent. However, as it removes the incentive to bring forward investment in the near-term relative to the temporary policy, a mildly negative and temporary demand effect reduces real GDP by 0.1 per cent in 2024-25.

Chart A: Real GDP impacts

Longer-term impacts on supply beyond 2028-29

Some Autumn Statement policies might also be expected to have significant effects beyond our five-year forecast. The impact of permanent full expensing on potential output will continue to build, from 0.1 per cent in 2028-29 to slightly below 0.2 per cent in the long run, as the economy gradually moves to a larger long-run capital stock. By contrast, the decision to hold capital departmental spending flat in cash terms for one additional year, means public sector fixed capital formation falls as a share of GDP. It now falls from 3.6 per cent this year to 3.1 per cent in 2028-29, reducing the public capital stock as a share of GDP. If sustained, this would likely also have a material, negative impact on potential output beyond the forecast horizon, relative to maintaining the public capital stock as share of GDP.

Other measures could also boost business investment particularly over the longer-term, including the proposals announced in the Chancellor’s Mansion House speech, changes to the implementation of the Solvency II regime, and measures affecting electricity grids and planning, and the regulated utilities. For instance, were the Government to achieve its ambition to accelerate the consolidation of pension schemes, so enabling them to achieve greater economies of scale, this could incentivise additional investment in infrastructure. It is also possible that some measures could provide a further boost to labour supply, such as proposed changes to fit notes. We have not explicitly adjusted our forecast for these changes, because of the significant uncertainties surrounding these impacts and the difficulty in assessing the extent to which regulatory changes are already implicit in our projections.

Impacts on inflation

The combined effect of changes to alcohol and hand-rolling tobacco duty reduces inflation slightly in 2024-25 and then increases it slightly in 2025-26. The small boost to aggregate demand from the government’s policy package, raises inflation in the near-term incrementally, increasing consumer prices by 0.1 per cent at our forecast horizon.

a) OBR, Dynamic scoring of policy measures in OBR forecasts, November 2023.

b) We review these estimates periodically. See our 2019 Forecast evaluation report and November 2020 Economic and fiscal outlook.

c) This figure is informed by estimates and insights from a range of studies, including those outlined in Meghir, C., and D. Phillips, Labour Supply and Taxes in Dimensions of Tax Design: Volume 1 of The Mirrlees Review, 2010, those in Adam, S., and D. Phillips., An ex-ante analysis of the effects of the UK Government’s welfare reforms on labour supply in Wales, 2013, and Attanasio, O., P. Levell, H. Low, and V. Sanchez-Marcos, Aggregating Elasticities: Intensive and Extensive Margins of Female Labour Supply, 2015. Professor Hamish Low provided useful guidance on this empirical literature.

Inflation

2.10 Inflation has fallen from its 41-year high of 11.1 per cent in October 2022, but not as sharply as we expected in our March forecast. In the third quarter of 2023, CPI inflation fell to 6.7 per cent, which was 1.3 percentage points higher than in our March forecast. We think three key factors explain why inflation was higher than our forecast. The first is a stronger pass-through of costs onto other prices. In Box 2.1 of our October 2023 Forecast evaluation report, we concluded that the knock-on effect of higher energy prices on wider costs was just under 50 per cent. By contrast, we had previously assumed a 25 per cent pass-through since our March 2022 forecast.^[4]
The second key factor is a degree of excess demand, as the economy seems recently to have been running slightly further ahead of potential rather than demand falling short of it as we had previously expected (see paragraph 2.17). The final factor, related to the second factor, has been faster nominal wage growth, as the labour market has proved more resilient than we expected in March (see paragraph 2.35).

2.11 We now expect inflation to fall more gradually over the next few years. Our central forecast sees CPI inflation at around 4.8 per cent in the final quarter of 2023, 1.9 percentage points above the March profile.^[5]
Domestic factors drive most of the upward revision, particularly higher nominal earnings growth outweighing the effect of lower energy prices. We forecast inflation to hit the 2 per cent target in the second quarter of 2025, about a year later than we forecast in March. It then dips below target in the middle of the forecast, driven by further falls in energy costs, along with increasing spare capacity in the economy, which reduces inflationary pressure on non-tradables (Chart 2.4). Inflation then returns to target by the forecast horizon.

Chart 2.4: Contributions to CPI inflation

2.12 Risks around the outlook for inflation remain high, given both domestic and international uncertainty. Our fan chart of potential inflation outcomes is normally constructed using historic forecast errors since 2003 (left panel Chart 2.5). This method suggests fairly contained risks to our projections, with only a 1 in 3 chance that inflation will be above 3 per cent in 2025. But this method understates the risks to our current forecast. Since 2020, the economy has undergone a series of major shocks from the pandemic and then the Russian invasion of Ukraine. The right-hand panel of Chart 2.5 shows how some of the forecast errors due to these shocks fell well outside our historic distribution. As discussed in paragraph 2.7, the current conflict in the Middle East poses a significant risk to the outlook for global energy prices. And more upward surprises to wage settlements represent a key domestic risk to the inflation outlook.

Chart 2.5: CPI inflation fan chart and past CPI forecast errors

2.13 RPI inflation has also fallen sharply from its peak but remains higher than we expected in our March forecast, despite a smaller-than-expected wedge above CPI. RPI inflation eased from 13.9 per cent in the final quarter of 2022 to 9.0 per cent in the third quarter of 2023, 0.9 percentage points higher than expected in March. Since the start of the year, RPI inflation has fallen by more than CPI due to differences in weights and coverage between both inflation measures – the most significant of which occur in housing. In our central forecast, RPI inflation is expected to fall to 6.7 per cent by the final quarter of 2023. We then expect it to fall further in the next two years to 5.1 per cent in 2024 and 2.6 per cent in 2025 as CPI inflation slows, house prices fall and mortgage rates ease.

2.14 The GDP deflator, which measures the price of all domestically produced goods and services, saw an even larger overshoot relative to our March forecast than CPI and RPI inflation. Annual growth in the GDP deflator was 7.9 per cent in the second quarter of 2023, 4.4 percentage points higher than our March forecast (see Chart 2.6). The change was spread across components, with higher-than-expected private consumer expenditure inflation, a higher implied price of government output, and a faster recovery in the terms of trade.^[6]
We expect the GDP deflator to average 7.0 per cent growth over 2023. We then forecast a rapid easing in inflation in 2024, as private consumption inflation moderates, and the terms of trade stabilise. Thereafter, the forecast growth profile is broadly in line with CPI inflation. In Box 2.2, we explore the changes in the composition of inflation between our March and November forecasts, and their implications for the public finances. And in Chapter 5, we present some illustrative alternative inflation scenarios to discuss their different fiscal consequences.

Chart 2.6: Contributions to GDP deflator growth

Box 2.2: The source of inflation and its fiscal consequences

While it has fallen significantly over the course of 2023, CPI inflation is higher, and is expected to remain above the 2 per cent target for longer, than we had assumed in previous forecasts, as shown in the top left panel of Chart B. Our March 2023 central forecast was for inflation at 6.1 per cent in 2023 (well above the roughly 2 per cent a year in our October 2021 forecast), before falling sharply to average just under 0.8 per cent between 2024 and 2027. In this forecast, it averages 7.5 per cent over 2023 and then falls more slowly, averaging 2.1 per cent between 2024 and 2028. The CPI is therefore 7.0 per cent higher by the start of 2028 than in March (top-right panel of Chart B).

As well as being more persistent, we now expect inflation to be more domestically generated than in our March 2023 forecast, rather than being driven largely by external factors. This change in the expected composition of inflation is due to:

• Lower energy prices. Our March inflation forecast was driven by a sharp increase in the cost of imported energy. As shown in the bottom left panel of Chart B, gas prices are now expected to be around 36 per cent lower by the end of 2023 than in our March forecast. By 2027-28, the gap is still 26 per cent. This puts downward pressure on ‘imported’ energy inflation;
• Stronger second-round effects from the energy shock. We now estimate that the knock-on effect of energy costs on wider prices has been almost 50 per cent of their direct contribution to CPI inflation (rather than the 25 per cent assumed since our March 2022 forecast).^a So, this pushes up ‘domestic’ inflation by more than in our previous forecasts;
• Less economic slack relative to March. As explained in Paragraph 2.17, we judge that the economy was operating around 0.5 per cent above its capacity in the second quarter of 2023. This is well ahead of the 1.3 per cent negative output gap we expected in March. A higher degree of excess demand in the first years of the forecast drives earnings and profits higher. Compared to March, we expect nominal earnings to be up 1.3 per cent by end-2023 and up 3.8 per cent by 2027-28 (bottom right of Chart B).

Chart B: Changes to inflation and inflationary pressures

A sharper fall in imported energy prices, stronger growth in wages and profits, and revisions to outturn data have delivered a faster-than-expected recovery in the UK’s terms of trade (the price of UK exports relative to imports). We now forecast the terms of trade only slightly below levels seen prior to Russia’s invasion of Ukraine over the next five years (left panel, Chart C). This stronger recovery means more of the rise in CPI inflation is reflected in higher GDP deflator growth (a measure of the price of domestically-produced goods and services).^b In March, we expected CPI and RPI inflation of 6.1 and 8.9 per cent respectively in 2023, but GDP deflator growth of just 2.9 per cent. We now expect 1.3 and 1.1 percentage points higher CPI and RPI inflation, respectively, at 7.5 and 10 per cent in 2023. But in an even larger revision, we expect 4.1 percentage points higher deflator growth, averaging 7.0 per cent in 2023 (right panel).

Chart C: Terms of trade forecasts and inflation in 2023

This shift in both the level and composition of inflation over the forecast period has important implications for the public finances. This is because:

• The GDP deflator is a key driver of tax revenues, whose bases (including wages, consumption and profits) depend on the nominal value of domestic economic activity. Frozen personal tax thresholds deliver a further boost to receipts through fiscal drag – as stronger nominal wage growth pushes more taxpayers into higher tax bands.
• CPI and RPI are key drivers of the cost of welfare benefits (most of which are indexed to CPI) and debt interest spending. Around a quarter of public debt is indexed to RPI and higher CPI inflation expectations tend to push up nominal interest rates on other government liabilities such as conventional gilts and Bank of England reserves.
• Departmental expenditure limits (DELs) over the current Spending Review period (up to 2024-25) are eroded by higher inflation, despite an increase in cash DELs at this Autumn Statement (£5.9 billion on average). From 2025-26, the Government assumes that day-to-day spending grows by an average of 0.9 per cent a year in real terms (using the GDP deflator), and capital spending remains frozen in cash terms. As a result, higher inflation does not fully feed through to higher departmental spending totals, and instead increases the cost pressure on departmental budgets within these overall totals.

These effects mean that this higher and more domestically driven inflation explains much of the improvement since March in the pre-measures fiscal outlook. It has led to a significant boost to nominal receipts, only partly offset by higher welfare and debt interest spending. Given departmental spending has not been fully adjusted to reflect inflation, it also leads to a decline in its real value. These fiscal consequences are set out in detail in Chapter 4.

a) For more information, see Box 2.1 of our 2021 Forecast evaluation report.

b) This is because CPI and RPI measure the prices of a representative basket of goods and services consumed by households in the economy, including imports, while the GDP deflator only captures prices for domestically produced goods and services.

Historical GDP revisions and the starting output gap

2.15 Historical GDP revisions suggest that the economy recovered more quickly and more fully from the pandemic than previously estimated. In the Quarterly National Accounts consistent with Blue Book 2023, the ONS revised up GDP growth by 0.6 percentage points in 2020 and 1.1 percentage points in 2021. There were minimal revisions to growth rates in 2022 and 2023. The revised data now suggest that output reached its pre-Covid level in late 2021 (which we had not expected to occur until 2024 back in our March forecast). Real GDP growth since March has also been stronger than expected, which leaves the starting point for real GDP 2.9 per cent higher in the second quarter of 2023 than we assumed in our March forecast and 1.8 per cent above its pre-pandemic level.

2.16 Stronger domestic inflation pressures indicate that some of this higher output reflects demand running somewhat ahead of the economy’s supply-side capacity. Our suite of output gap models also suggest that demand exceeded capacity in mid-2023, particularly in the labour market. Job vacancies were 1.5 times higher than pre-pandemic levels in the third quarter of 2023 and indicators of recruitment difficulties remain above historical averages. Business surveys of capacity utilisation were close to normal levels.^[7]

2.17 We therefore judge that the economy was operating around 0.5 per cent above capacity in the second quarter of 2023, as opposed to expecting the economy to be operating 1.3 per cent below capacity in the March forecast. As a result, we judge that the level of potential output at the start of the forecast is around 1.0 per cent higher than anticipated back in March, though the level of actual GDP is 2.9 per cent higher.

Potential output forecast

2.18 We now expect potential output to grow by an average of 1.6 per cent a year between 2024 and 2028, 0.1 percentage points slower than in our March forecast. The drivers of these changes are described further in later sections. In summary, (left-hand panel, Chart 2.7):

• Labour supply contributes 0.7 percentage points to potential output growth, unchanged from March. Higher migration and the impact of Autumn Statement policies together boost labour supply, but are offset by a gradual fall in average hours worked (see Paragraph 2.20).
• Capital deepening contributes 0.2 percentage points, around 0.1 percentage points less than in March. The recent strength in business investment and the announcement of permanent full expensing both boost the capital stock. But their impact is offset by our revised assumption of a rising, rather than constant, retirement rate of capital (see Paragraph 2.23).
• The contribution of total factor productivity (TFP) grows through the forecast period and averages 0.7 percentage points a year, which is 0.1 lower than March. The weaker growth is concentrated at the start of the forecast due to historical revisions to the level of real GDP that suggest more of the expected post-pandemic rebound in productivity has already occurred (see Paragraph 2.25).

2.19 Revisions to historical data and stronger-than-expected recent outturns have raised our estimate of the level of potential output at the start of the forecast by 1 per cent. Combined with the slightly weaker outlook for potential output growth in this forecast, that leaves the level of potential output at the end of March’s forecast period in 2027 0.5 per cent higher than our March estimate (right hand panel of Chart 2.7). The net contribution of policies in this Autumn Statement is to increase potential output by 0.3 per cent by 2028-29. Some of the boost to potential output from policy measures builds further in years beyond the forecast horizon (see Box 2.1).

Chart 2.7: Decomposition of changes to potential output

Labour supply

2.20 Labour supply growth of 0.7 per cent a year on average between 2024 and 2028 is broadly in line with population growth. We define labour supply as the equilibrium total hours worked across the economy. A gentle rise in the trend participation rate is broadly offset by a slight fall in trend average hours worked. The level of labour supply in 2027 is 0.3 per cent lower than in our March forecast due to the net effect of three changes:

• The 16+ population is expected to be around 70,000 (0.1 per cent) smaller by 2027 than we forecast in March, as a lower starting population is only partly offset by a higher migration forecast in the near term.^[8] The change in our migration assumption raises our average forecast for total population growth between 2024 and 2025 by 0.1 percentage points to 0.8 per cent a year. Population growth then slows back to a similar rate to the March forecast by 2027. Box 2.3 discusses the reasons for the upward revision to our net migration assumption.
• Stronger outturn data, part of which we assume is structural, and the labour supply policies announced in this Autumn Statement increase our medium-term trend labour participation rate forecast by 0.3 percentage points in 2027 compared to March. Policy measures (discussed in Box 2.1) increase participation by 81,000 in 2028-29, accounting for 0.2 percentage points of the upward revision. Despite these improvements, the trend participation rate remains 0.6 percentage points below its pre-pandemic peak of 63.8 per cent in 2019.
• Our forecast for trend average hours worked is 0.6 per cent lower by 2027 than in March, as a reassessment of the effect of demographic trends outweigh the announced cut to National Insurance Contributions (NICs). Projected demographic trends point to a workforce with a smaller share of individuals aged 25-64 and a larger share of workers aged either 16-24 or over 65 between 2023 and 2028. Following modelling work over the summer, we now assume this trend will reduce the average weekly hours worked to 31.7 in 2027. This is 0.2 hours a week below the March forecast which lowers potential output by 0.5 per cent. This change is only partly offset by the 0.1 per cent boost to average hours provided by the NICs cut.

Box 2.3: Net migration and its impact on labour supply

Net migration has increased substantially in the wake of the pandemic, mainly driven by an increase in humanitarian visas, international students, and dependants. The latest ONS data show net migration was more than 600,000 in 2022,^a compared to around 220,000 in 2019. This was entirely driven by an increase in non-EU immigration. EU net migration has turned negative under the new post-Brexit immigration regime and stood at -50,000 in 2022. Home Office data show visas granted continued rising in the first half of 2023, on a rolling annual basis, with humanitarian visa issuance falling but the remaining categories continuing to rise (left panel, Chart D). In terms of sub-categories, visas for health and care workers showed a substantial pickup, growing by more than 150 per cent in the year to June 2023.

Future migration levels are highly uncertain and difficult to forecast, but high recent immigration levels should also result in higher levels of emigration in the coming years. This is especially likely for students, as most tend to leave a few years after their arrival when they have completed their studies. However, international students are now likely to stay for longer, as the new Graduate Visa route allows them to stay for 2 years after their graduation. The medium-term stay rate of (non-temporary) workers has also historically been fairly low, though the recent inflow of migrants come from countries and work visa subtypes that have traditionally had higher ‘stay rates’. Accounting for those changes in the composition of migration, Hall et al (2023) estimate workers’ average stay rate will increase from 25 per cent between 2010 and 2014 to 57 per cent under the new regime. A summary of historical ‘stay rates’ for some visa groups is shown in the right panel of Chart D.

Chart D: Visas granted by category and migrants’ stay rates 8 years after arrival

This likely increase in emigration, alongside recent changes in government policy, means that we assume net migration will gradually fall from current high levels. In the absence of updated population projections, we retained the ONS’s 2020-based projections for long-term net migration settling at 245,000,^b which is also broadly consistent with the Government’s ambition to bring net migration back to pre-pandemic levels. Since March, the Government has taken several measures to tighten the migration regime including the increase in visa fees, the immigration health surcharge and restrictions on student dependants, which support this trajectory. Taking account of the impact of government policy as well as recent studies of post-Brexit migration flows,^c we have assumed a transition path from the higher recent migration outturns to the ONS’s long-term population projection from 2025-26. This results in a higher near-term path for net migration than forecast in March (left panel, Chart E) and adds 150,000 more people to the population over the next five years.

To forecast the impact of this higher near-term net migration on total labour supply, we assume that migrants have the same labour market participation rate as the UK resident population. This is consistent with ONS time-series data on the participation rates of different migrant groups and the overall composition of recent arrivals under the post-Brexit migration regime. The LFS suggests that the average participation rate of the UK adult population was around 63 per cent in the year to June 2023, while that of recently arrived migrants was 67 per cent (right panel, Chart E). Those migrants who arrive for work reasons tend to have the highest participation rates of over 90 per cent. Those that arrive as students, dependants, or for other reasons tend to have below average labour participation rates, albeit with some convergence in recent years. As the LFS sampling is likely under-representing students and dependants, who have lower participation rates than workers, the average participation rate for all migrants is probably an overestimate. Over time, this may be partly offset by migrants, on average, becoming increasingly likely to participate in the labour market the longer they stay and settle in the UK.

Chart E: Net migration and the participation rate of recent migrants by reason

Based on these assumptions, the increase in our migration forecast relative to March adds 70,000 people to the labour force over the next five years. While these assumptions are highly uncertain, we judge the risks around them to be balanced. Future net migration levels could be higher than assumed in our forecast, if work or student visa issuance remains high or if a larger share of new migrants end up staying in the UK. On the other hand, recent humanitarian arrivals from Ukraine, Hong Kong, and other countries could leave in greater numbers, either to return home or move to a third country. Similarly, there is considerable uncertainty around the participation rate of inward migrants, which could be higher or lower than we have assumed. We will continue to review these judgements in light of upcoming ONS data releases on population and the labour market.

a) ONS international migration data for 2022 are provisional estimates informed by past behaviour and are subject to revisions.

b) ONS’s 2020-based interim national population projections account for international migration data up to year ending June 2022.

c) Hall, T., A. Manning and M. Sumption, Projecting UK net migration, Centre for Economic Performance, Occasional Paper No.60, October 2023.

Investment and the capital stock

2.21 Relative to March, the real capital stock per hour worked is unchanged by 2027, leaving our forecast for the total contribution of capital deepening to potential output over the forecast period the same. This reflects a higher forecast for business investment being fully offset by a higher assumed retirement rate of capital.

2.22 In terms of the drivers of business investment over the forecast period:

• Upward revisions to the historical levels of business investment over 2020-2022 mean capital spending has recovered more completely from the pandemic than previously indicated. Cumulative business investment over this period was 3.6 per cent higher than in the data available back in March.
• We expect a fall in business investment in the near term as the economy moves from a period of excess demand to a period of excess supply and as past increases in interest rates raise the cost of capital and weigh on capital spending. We then expect a slight rise as interest rates fall back and as excess supply in the economy is eroded. Investment grows by less over the forecast than in March as the historical revisions mean that there is less scope for further cyclical recovery and as the unexpectedly large rise in interest rates since March raises the cost of capital.
• As discussed in Box 2.4, the decision in this Autumn Statement to make 100 per cent capital allowances a permanent measure lowers near-term investment, but boosts it in the medium term and beyond. This raises cumulative investment by 0.3 per cent in 2027 relative to March.
• The overall effect of these changes is that cumulative business investment between 2023 to 2027 is expected to be 4.5 per cent higher than in March. This is because the higher starting point is only partly offset by slower growth in business investment over the forecast period.

2.23 The effect of this increase in gross investment on the stock of capital is offset by an increase in the assumed retirement rate of existing assets. The retirement rate of capital is the pace at which capital assets become obsolete, are withdrawn, and need to be replaced. This rate has risen steadily from 3.7 per cent in 2000 to 4.7 per cent in 2022, reflecting a growing share of shorter-lived intangible assets, such as computer software, in the capital stock. Having previously assumed that the retirement rate would stabilise and remain constant over our forecast period, we now assume intangible assets continue to grow as a share of the capital stock. So, the capital retirement rate continues rising to 5.0 per cent by 2028, reducing the capital stock by 0.7 per cent in 2027 compared to holding the rate flat.

Box 2.4: The impact of corporation tax changes on business investment

The last three years have witnessed four major changes in the general corporate tax system:^a

• The March 2021 Budget announced a permanent increase in the headline rate of corporation tax from 19 to 25 per cent for firms with profits over £250,000. This took effect from April 2023. The 19 per cent tax rate was retained for smaller firms, which account for 18.5 per cent of total receipts.
• The same Budget introduced a temporary 130 per cent super-deduction for all qualifying plant and machinery investments undertaken between April 2021 and March 2023.
• In the March 2023 Budget, the Chancellor announced temporary full expensing, which gave a 100 per cent tax deduction for all qualifying plant and machinery investments undertaken between April 2023 and March 2026.
• In this Autumn Statement, the Chancellor has announced permanent full expensing for all qualifying plant and machinery investments.

This succession of tax changes took place against the background of a rapidly changing economic landscape, including supply bottlenecks for key inputs, a spike in inflation, and a dramatic rise in the cost of capital from higher interest rates. Against this volatile background, this box considers the evidence on how business investment has responded to past tax changes and the likely impact of permanent full expensing.

Corporation tax increase and the super-deduction

The increase in the corporation tax rate was previously assumed to reduce the optimal long-run capital stock and business investment via a higher long-run cost of capital. In isolation, we estimated this would lower potential output by 0.2 per cent over the forecast period, increasing to 0.3 per cent in the longer term.

In 2021-22 and 2022-23, we assumed firms would bring forward some investment to take advantage of the temporary 130 per cent super-deduction, which we initially assumed would have, on its own, a peak business investment impact of 10 per cent. As well as boosting investment in its own right, this measure provided a counteracting effect to the incentive firms would have otherwise had to delay investments to take advantage of capital allowances being more generous after the corporation tax rate rise in April 2023.^b

We subsequently revised down the impact to 5 per cent in March 2022, informed by survey evidence from Deloitte and the CBI suggesting lower take-up and strong corporation tax data. Based on the latest outturn data, there is some evidence that the super-deduction brought forward some qualifying investments, although it does not seem to have significantly boosted overall business investment. Over the two-year super-deduction period, business investment has risen for some asset groups that were eligible for the super-deduction, such as transport and ICT equipment and machinery (Chart F). However, classifications of qualifying assets and data for business investment asset groups do not fully align, making it difficult to draw firm conclusions.^c

Chart F: Business investment by asset type

More recent survey evidence has indicated a smaller but still positive impact from the super-deduction, with the Business Insights and Conditions Survey (BICs) suggesting at least a 2 to 3 per cent increase in business investment from the policy for a ‘representative’ firm – lower than our estimated peak of 5 per cent. This may reflect the fact that large firms who were more likely than others to use the super-deduction, typically invest more, or that they responded more to the changes in their incentives than a representative firm. It could also be due to the limited time window for businesses to take advantage of the subsidy. Overall, given the disruptions from the pandemic, energy crisis, and Brexit, and the impact of the pre-announced rise in corporation tax, the fact that investment did not fall over the past two years suggests that the super-deduction worked as intended to support investment.

Temporary and permanent full expensing

At the time of our March 2023 forecast, we expected the new temporary investment allowances announced in the Budget to incentivise firms to bring forward further capital spending in the qualifying timeframe of April 2023 to March 2026. This boosted investment by just over 3 per cent a year over a period of three years (as the scheme’s longer duration meant the incentive to concentrate investment was weaker than we assumed for the years the super-deduction was in place). So far, respondents to the Bank of England’s latest Decision Maker Panel survey have reported an average expected rise in investment of 1.5 per cent over the years that temporary full expensing was in place.^d

In this Autumn Statement, the Chancellor has announced that the 100 per cent capital allowances measure will be made permanent. By delivering a lasting reduction in the cost of capital, we assume this would induce a rise in the economy’s optimal capital stock of 0.5 per cent in the long run, which was absent in the case of temporary capital allowances. Permanent full expensing removes the near-term incentive to bring investment forward into the period over which the temporary policy would have been in place. This means that real investment is lower in the first half of the forecast period by £11 billion compared to what we previously assumed (Chart G). But it is higher in the medium term by £25 billion. Overall, the shift from temporary to full expensing has increased total business investment by £14 billion over the forecast period or £3 billion a year (1.2 per cent on average). For discussion of the GDP implications of this higher investment, see Box 2.1.

Chart G: Business investment forecast

a) This excludes multiple, smaller changes like taxes on energy companies and the plethora of smaller changes to tax reliefs.

b) This would have incentivised some firms to delay investment until the new rate came into force in April 2023. Firms could then make deductions at a corporation tax rate of 25 per cent rather than 19 per cent as noted in Chapter 6: Corporation tax and investment of the Institute for Fiscal Studies’ (IFS) October 2022 Green Budget.

c) For example, the Institute for Fiscal Studies (IFS) points out differences between recent data for ICT and transport equipment and the classification of plant and machinery for tax purposes in Chapter 10: Full expensing and the corporation tax base of their October 2023 Green Budget.

d) See the Bank of England’s October 2023 Monthly Decision Maker Panel data for more.

Total factor productivity

2.24 Total factor productivity growth rises from recent subdued rates to reach around 0.9 per cent in 2028, above the average seen in the decade after the financial crisis but still below the rates seen before the financial crisis. This gives an average contribution to potential output growth from 2023 to 2028 of 0.7 percentage points. The outlook for TFP and overall productivity is extremely uncertain and we explore alternative scenarios where productivity growth is stronger and weaker than in our central forecast in Chapter 5.

2.25 The level of TFP in 2027 is 0.8 per cent points higher than in March. This reflects the effects of three partly offsetting shifts:

• Blue Book revisions to historical GDP growth in 2020 and 2021 lifted output by more than the effect of higher investment on capital deepening with an unchanged estimate of the amount of hours worked. This suggests that TFP made a more complete recovery from the pandemic than suggested by data available in March.
• But the very latest data suggests that post-pandemic growth in TFP has been slower than we had expected in March. Reflecting this loss of momentum and a judgement that part of the expected recovery from the pandemic has already happened, we have lowered average TFP growth over 2023 and 2024 by 0.4 percentage points.
• Lower medium-term gas prices reduce economy-wide unit energy costs and provide a 0.2 per cent point boost to the level of TFP by the end of the forecast relative to March.^[9]

Real GDP and the output gap forecast

2.26 While real GDP growth has been stronger so far this year than we anticipated in March, we expect it to cool over the second half of 2023. Quarterly real GDP growth was 0.3 per cent in the first quarter of 2023 and 0.2 per cent in the second quarter whereas, in March, we expected output to contract by 0.4 per cent over the two quarters. But we think that this unexpected strength will be reversed in the coming quarters. This is because interest rates have risen by more than expected in March which should weigh on activity. Business surveys and consumer confidence measures are also consistent with weak near-term growth. Therefore, we forecast GDP to remain flat in the third quarter of 2023 and only grow by 0.1 per cent in the fourth quarter, below the combined growth of 0.7 per cent we expected in March. This results in an annual growth forecast of only 0.6 per cent in 2023.

2.27 We expect growth to remain subdued at 0.7 per cent in 2024 as a result of weak real wage growth, the effect of past increases in interest rates and fading fiscal support weighing on economic activity. It is plausible that only around half the impact of the past substantial rise in interest rates has yet fed through to the economy, with the rising share of fixed rate mortgages in recent years slowing the pass-through from mortgage rates to household disposable incomes.^[10]
In our central forecast, growth picks up to 1.4 per cent in 2025 and 2.0 per cent in 2026 and 2027 as inflation falls back from recent elevated rates and provides a further boost to real wage growth. A fading in the effect of past interest rate increases and the outright falls in rates in our conditioning assumptions also push up on GDP growth relative to supply growth. Growth then drops back towards our new estimate of medium-term potential output growth of 1.6 per cent in 2028.

2.28 Real GDP growth is forecast to average 1.5 per cent between 2024 and 2027, 0.6 percentage points weaker than forecast in March. The lower growth profile is due to a higher estimated starting output gap providing less scope for above-potential growth than in March (paragraph 2.17), a weaker forecast for potential output growth (paragraph 2.18) and a later closing of the output gap pushing some of the recovery beyond the March forecast horizon (see below). However, the level of real GDP is slightly higher in 2027 than in March as the weaker growth forecast only partly offsets the upward revisions to historical GDP data (see Chart 2.8).

Chart 2.8: Real GDP forecast

2.29 We expect the economy to move from the current modest amount of excess demand (outlined in paragraph 2.17) towards a modest degree of excess supply, with output gap bottoming out at -0.9 per cent of potential output in the middle of 2025. This reflects continued subdued real GDP growth in the face of weak real wage growth, higher interest rates and fading fiscal support. As these factors fade, GDP growth picks up to rates above potential output growth and the negative output gap closes by the forecast horizon. The trough in the output gap is 0.7 percentage points higher than March and around two years later, reflecting the unexpected resilience of the economy in the face of the recent energy price shock, counterbalanced by interest rates rising by more than expected. Policy measures in the Autumn Statement that boost demand are expected to lift the trough in the output gap by 0.1 percentage points relative to our pre-measures forecast.

Chart 2.9: Output gap forecast

2.30 There is significant uncertainty around our GDP growth forecast. The left-hand side of Chart 2.10 illustrates our central forecast for real GDP growth within a fan chart. The shaded areas illustrate the probability of different growth outcomes, conditioned on the size of past forecast errors since 1988. These uncertainty bands imply that there is around a 1 in 4 chance that real GDP growth will be negative in the latter half of the forecast and a similar probability that growth will exceed 4 per cent. But, as with our inflation fan chart, this likely understates the uncertainty around our current forecast give the scale and array of risks we currently face. The right-hand panel of Chart 2.10 shows how many of the recent errors around our forecast fall well outside our historic distribution, owing to the shocks of the pandemic and energy crisis. An uncertain outlook for energy prices, inflation, and productivity growth pose particular risks to our forecast. Chapter 5 presents scenarios showing how different paths for inflation and real output growth affect the public finances.

Chart 2.10: Real GDP growth fan chart and past GDP forecast errors

Labour market

Participation

2.31 Following the pandemic, the number of working-age people classed as inactive jumped by almost 650,000 at its peak, and it remains 400,000 above pre-pandemic levels. We discussed the different causes behind this sharp rise in inactivity in detail in Chapter 2 of the 2023 Fiscal risks and sustainability report. The contribution of different groups to the rise is shown in the right-hand panel of Chart 2.11 and our assessment of recent trends and the outlook for them is that:

• The rise in inactivity has been more than accounted for by those citing long-term sickness as the reason and is likely to prove more long-lasting than other types of inactivity. We expect long-term health conditions to continue to weigh on participation, which contributes to the rise in our incapacity benefit caseload forecast discussed in Chapter 4. The welfare and other reform measures announced by the Chancellor alongside this Autumn Statement should provide a partial offset to this trend, raising participation by around 50,000.
• The effects of an ageing population that weighs on participation will intensify in the coming years,^[11] due to the increasing share of older people in the population who have lower-than-average participation rates.
• Student inactivity rose significantly during the pandemic but has fallen back more recently and has been one of the main drivers of the unexpected rise in participation since our March forecast. Surveys suggest that some have joined the labour market due to cost-of-living pressures.^[12] We therefore expect that some of the recent fall in student inactivity is cyclical rather than structural and that this will begin to unwind as real incomes resume growing.
• We anticipate a fall in inactivity due to caring responsibilities as recent drops in birth rates mean that the share of children in the population is set to fall in the medium term. We estimated that the expansion of childcare provision announced in the Spring 2023 Budget will also boost participation by 60,000 people by 2027-28.

2.32 The combination of these partly offsetting trends results in a forecast for the overall participation rate that falls back in the near-term, before rising marginally in the medium term (left hand panel of Chart 2.11. The welfare reforms announced alongside this Autumn Statement, and the reduction in national insurance contributions, are expected to provide a boost of around 0.2 per cent to participation by 2028-29 (see Box 2.1 for more details).

Chart 2.11: Participation forecast and inactivity by reason

Unemployment and employment

2.33 We expect the labour market to continue loosening with the unemployment rate peaking at 4.6 per cent in the second quarter of 2025, as GDP growth slows and spare capacity opens up. Labour demand has been weakening recently, with vacancies falling from a peak of 1.3 million in May 2022, to around 960,000 in October 2023. The employment rate is forecast to fall from 60.7 per cent in the third quarter of 2023 to 60.2 per cent in the second quarter of 2025, reflecting both rising unemployment and falling participation. The employment rate then makes a partial recovery to 60.6 per cent by the end of the forecast, 0.3 percentage points higher than we expected in our March forecast.

Chart 2.12: Unemployment and employment rates

2.34 Since October, updates of the Labour Force Survey (LFS), which is the primary source of our population and labour market data, have been temporarily paused as the ONS undertakes improvements to address issues with falling sample sizes.^[13]
More timely, alternative, estimates of population and migration suggest that the total number of adults in employment, unemployment and inactivity are higher than the LFS estimates.^[14] In addition, recent Bank of England analysis indicates that because the LFS has not captured the ageing of the population since mid-2021, it may be overestimating employment and participation rates by around 0.4 percentage points, but the unemployment rate is likely less affected.^[15] The net effect of these issues is hard to gauge and makes it more difficult to assess the starting point for our labour market forecast.

Average earnings

2.35 Our forecast for average earnings growth is around 2 percentage points higher in both 2023 and 2024 than in March. Whole economy earnings growth hit just over 8 per cent in the second quarter of 2023 and ranged between 6 and 8 per cent in the third quarter, according to official statistics.^[16] This was driven by strong private and public sector pay growth in recent months, although the latter was also boosted by the timing of one-off lump sum payments.^[17] High average pay outturns, stronger than expected inflation and a still-tight labour market conditions mean we now expect average earnings growth of 6.8 per cent across 2023, 1.9 percentage points higher than our March forecast. We expect earnings growth to ease back to 3.7 per cent in 2024 and 2.2 per cent in 2025 as inflation falls further, labour market conditions continue to loosen and unemployment rises. Nominal earnings growth is then broadly steady until 2028, when it rises to 2.8 per cent as the output gap closes.

Chart 2.13: Average earnings

Composition of economic activity

Credit conditions and the housing market

2.36 Average interest rates on the stock of mortgages are expected to rise from a low of 2 per cent in 2021 to a peak of 5 per cent in 2027. This is 0.8 percentage points above our March forecast and 2.2 percentage points above the average of the previous decade. The Bank’s November Monetary Policy Report noted that higher mortgage rates are likely to take longer to pass through to the stock of mortgages than in the past. Prior to the financial crisis, when Bank Rate was last raised to current levels, only around half of all outstanding mortgages were fixed rate contracts. Today around 85 per cent are on a fixed term, with two thirds of these on a term longer than two years.

2.37 In the second quarter of 2023, housing transactions fell to their lowest level since the middle of the pandemic, as higher mortgage rates reduced housing affordability. Leading indicators suggest the market will remain weak, for example, in October RICS reported that new buyer enquiries are at their lowest level (outside the pandemic period) since 2008. Residential property transactions are expected to fall as the housing market continues to cool. We expect housing transactions to fall by 6.9 per cent in 2024, a 1.9 percentage point steeper decline than in our March forecast of 5 per cent. We then expect housing transactions to steadily return to growth from the final quarter of 2024, returning to pre-pandemic levels in the first quarter of 2027.

2.38 Our central forecast estimates that house prices will grow by 0.9 per cent in 2023 and then fall by 4.7 per cent in 2024. This would be consistent with the price of the average UK home reaching a low of around £266,000 at its trough in the final quarter of 2024. All in all, from their high in the fourth quarter of 2022 to their low in the final quarter of 2024, nominal house prices are expected to decline by 7.6 per cent (2.4 percentage points less than we expected in March).^[18] We then expect house prices to recover slowly, reaching their late 2022 peak levels in the second half of 2027 and rising to 6.4 per cent above this level by the end of the forecast. The outlook for house prices is particularly sensitive to changes in interest rates and household income growth.

Chart 2.14: Mortgage rates and house prices

Household income

2.39 Real household disposable income (RHDI) has proved unexpectedly resilient since our March forecast (as discussed in detail in Box 2.2 of our latest Forecast evaluation report). In 2022, RHDI fell by 1.4 per cent, a 1.1 percentage point smaller drop than estimated in March. We now forecast RHDI to rise by 0.6 per cent in 2023, partly because rising interest rates support household incomes (on aggregate) due to the boost to savings income from higher deposit rates so far outweighing the rise in interest payments from higher mortgage rates. But RHDI is expected to fall again in 2024, by 0.9 per cent, as inflation slightly outweighs growth in pay and non-labour income. The reduction in the growth contribution of non-labour income is because higher interest receipts are more than offset by a rise in debt interest payments, as more fixed-rate mortgages face renewal and banks recover their retail margins over deposit rates. RHDI gradually returns to growth of around 2 per cent in the medium term, as inflation drops back to around the 2 per cent target, pay growth reaches above-inflation rates and interest rate rises have fully passed through.

Chart 2.15: Contributions to real household disposable income growth

2.40 On a financial year basis, we expect RHDI per person, a measure of living standards, to fall by 0.8 per cent in 2023-24, following a fall of 2.2 per cent in 2022-23. This 3 per cent drop in living standards since 2021-22 is half the fall we forecast in March, as recent ONS outturns have surprised to the upside. While the trough in RHDI per person is shallower than in the March forecast, we expect its weakness to continue for longer, consistent with our downgrade to our forecast for near-term real GDP growth. And the overall 3.5 per cent peak-to-trough drop in RHDI per person between 2019-29 to 2024-25 is still the largest reduction in real living standards since ONS records began in the 1950s. By 2027-28, we forecast RHDI per person to be 0.9 per cent stronger than we expected in March, as higher prices and tax payments are outweighed by higher labour incomes and benefit receipts.

Chart 2.16: Real household disposable income per person

2.41 We expect the 2 pence cut to employee National Insurance contributions announced in the Autumn Statement to directly boost real household incomes per person by around 0.5 per cent by 2028-29. But the effect of the freezes in income tax and NICs thresholds – alongside recent strength in nominal earnings growth will result in a significant amount of ‘fiscal drag’ (described in detail in Box 3.1). We estimate that higher earnings have raised the direct cost to households from frozen tax thresholds relative to our March forecast by 0.7 per cent. Both policies should also have significant ‘indirect effects’ on household incomes via our economy forecast, including through their effects on labour supply. We have quantified the impact of the 2 pence cut to National Insurance contributions in Box 2.1, but not the effect of frozen tax thresholds given their more uncertain impact on work incentives. Nonetheless, it could have been a factor that has recently weighed on potential labour supply. We will return to this issue and potentially refine our judgement in future forecasts.

Household saving and consumption

2.42 The adjusted household saving rate is expected to fall from just above 3 per cent of incomes in 2022 and 2023 to 1.8 per cent in 2024, as consumers seek to maintain real spending despite falling real incomes.^[19]
The saving rate is, on average, around 2½ percentage points higher in 2023 and 2024 than our March expectation, as aggregate household incomes have proved more resilient than anticipated (as described in paragraph 2.39). From 2025, as inflation returns to target and income growth stabilises, the adjusted saving ratio rises to 2¾ per cent. The headline saving ratio, accounting for pension fund accumulation, is expected to remain just above 8 per cent in 2023, similar to the previous year, before settling around its historical average of just below 8 per cent over the rest of the forecast. There are substantial differences in saving patterns across the income and wealth distribution. Low-income households are expected to benefit relatively less from higher returns on savings and are more likely to be affected by rising borrowing costs. However, they also tend to adjust their spending more strongly in response to income changes.^[20]

2.43 While consumption has been more resilient than expected in the first half of 2023, we have lowered our forecast for household spending growth over 2024 and 2025. We expect consumption growth of 0.5 per cent in 2023, 1.3 percentage points higher than forecast in March. However, we have lowered our forecast for cumulative consumption growth in 2024 and 2025 by 1.7 percentage points to 1.6 per cent. This reflects a combination of more persistent weakness in household incomes and higher interest rates than expected in March. During this period, households are expected to support consumption by reducing saving. From 2026 to the end of our forecast, we expect consumption growth to average 1.9 per cent as interest rates stabilise, the unemployment rate declines, and RHDI recovers.

Trade and the current account

2.44 We expect trade volumes to stagnate in the medium term as weak economic growth weighs on the domestic economy and on the UK’s major export markets. Recent trade outturns have been volatile but remain consistent with our assumptions on the impact of Brexit. We continue to expect that exiting the EU will reduce the UK’s trade intensity (exports plus imports as a share of GDP) by 15 per cent in the long term. As of 2023, UK trade intensity remains 1.7 per cent below its 2019 level, versus an average increase of 1.9 per cent across other G7 economies (Chart 2.17). The relative weakness in UK trade since the pandemic is explained entirely by goods trade, as UK services trade has grown at a similar rate to other G7 countries. This may suggest that Brexit frictions and post-pandemic disruptions have weighed more on trade in goods than on services. From 2024 to 2027, we expect real exports to average growth of just 0.1 per cent a year, down from 0.5 in March. We expect real imports to decline by 0.4 per cent a year on average from 2024 to 2027, though this is up from 0.8 in March.

Chart 2.17: Trade intensity

2.45 The current account deficit widened in 2022, driven by surging prices for energy imports pushing the trade deficit wider after Russia’s invasion of Ukraine (Chart 2.18). As the shock spread, there was a rapid deterioration in the relative price of exports versus imports – the UK’s terms of trade. Yet 2023 has seen the terms of trade largely recover to 2021 levels. We expect the current account deficit to narrow from 3.7 per cent of GDP in 2023 to 2.4 per cent by 2028, driven mainly by the trade balance. As energy prices fall back further, the trade deficit is forecast to narrow from 1.4 per cent of GDP in 2023 to 0.3 per cent in 2028. After being in surplus in the previous two years, net investment income is estimated to return to deficit in 2023, as income earnt by overseas investors on their UK investments increases more than that earnt on the UK’s overseas investments. We forecast the deficit to be relatively flat over the rest of the forecast period.

Chart 2.18: Current account deficit

Sectoral net lending

2.46 Following large swings during the pandemic and energy crisis, sectoral balances are expected to return to historically more normal levels over the forecast period (Chart 2.19). As household saving falls, the household sector surplus declines in the early part of the forecast period from 1.9 per cent of GDP in 2023 to 1.2 per cent in 2024. The net lending position of households then strengthens by the end of 2024 as household incomes recover, settling at around 1½ per cent over the rest of the forecast. During the pandemic and energy crisis, fiscal support to the economy resulted in a large government deficit, balanced by large surpluses for households and, to a lesser extent, corporations. The corporate sector is expected to move from its recent net lending position to one of balance over the forecast period as investment rises. Mirroring the household and corporate sector, the government balance narrows over the forecast horizon as the fiscal deficit falls. And borrowing from the rest of the world gradually reduces, as the current account deficit narrows.

Chart 2.19: Sectoral net lending

Nominal GDP and its composition

2.47 Our forecast, together with revisions in historical data, leaves the level of nominal GDP 5.3 per cent higher in 2027-28 than we expected in March (a difference of £156 billion). The historical revisions largely stem from the 2023 Blue Book revisions to the National Accounts, which left the level of nominal GDP 2.1 per cent higher in 2022-23. The remaining 3.2 percentage points stem from faster growth thereafter, which is more than explained by a 5.3 percentage point upward revision to cumulative GDP deflator growth from 2022-23 (from the domestically generated inflation described in Box 2.2), which more than outweighs our weaker forecast for real GDP growth over the same period.

Chart 2.20: Nominal GDP

2.48 Breaking down these changes to nominal GDP since March into its income and expenditure components we see that:

• On the income side (left-hand panel), revisions are largely a result of rising labour incomes (which are more tax-rich than non-labour income). While little changed from our March forecast in 2022-23, upward revisions to labour income in subsequent years of the forecast explain an increasing share of the difference in nominal GDP. By 2027-28, labour income contributes 1.8 percentage points to the 5.3 per cent cumulative nominal GDP growth differential. Most of the remaining difference in nominal GDP relative to the March forecast is explained by ONS revisions to accounting items with minimal impact on activity or taxation. These include imputed rents of owner-occupiers and the impact of interest rate movements on the treatment of borrower financial services (FISIM)).
• On the expenditure side (right-hand panel), revisions are eventually dominated by the change in nominal consumption, itself driven by a higher consumption deflator (the price of consumption). Relative to our March forecast, nominal consumption is a little weaker in 2022-23 but contributes 3.9 per cent to the nominal GDP uplift in 2027-28 (this entirely reflects revisions to inflation, as real consumption’s contribution is 0.1 percentage point lower in 2027-28 relative to our March forecast). The contribution from corporate investment is driven by the timing effects of the extension to full expensing announced in the Autumn Statement. Other components of GDP have seen smaller changes and are less important for our fiscal forecast.

Chart 2.21: Nominal GDP: Changes since March

Comparisons with external forecasters

2.49 Our forecast for cumulative real GDP growth from 2023 to 2027 is 1.1 percentage points higher than the latest independent average forecast and 0.4 percentage points lower than the IMF’s October World Economic Outlook forecast for the UK.^[21]
The Bank of England expect much lower medium-term growth, which leaves the level of output 3.2 per cent lower than our central forecast, by the end of their forecast horizon in 2026. This difference is driven by several factors, including: a large projected negative output gap at the end of the Bank’s forecast period, different assumptions surrounding productivity growth, and small differences in their market determinants.

Chart 2.22: Comparison of forecasts for cumulative growth in real GDP

2.50 Table 2.2 compares our GDP growth, CPI inflation and unemployment rate projections to those produced by external forecasters. Our central forecast for CPI inflation in 2024 is 0.6 percentage points higher than the average of independent forecasters, but the same as the Bank’s latest forecast. In 2025 and 2026, our CPI forecast averages 0.3 percentage points lower than the Bank of England’s and the average of independent forecasters, reflecting our weaker wage growth forecast. Our unemployment projection for 2025 is 0.5 percentage points below the Bank’s 2026 forecast, which sees the unemployment rate continuing to climb through to 2026. Our unemployment forecast is 0.6 percentage points below the independent average forecast for 2027.

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

Chapter 3: Policy measures

Introduction

3.1 This chapter:

• sets out the main policy measures announced in this Autumn Statement and in the period since the March Budget, including how they have been incorporated in our forecast and the uncertainties around their costing;
• provides an update on selected previously announced measures; and
• discusses policy risks, which are measures or policy ambitions that are yet to impact our central forecast but have the potential to materialise.

Policy announcements in the Autumn Statement

3.2 This forecast incorporates the economic and fiscal implications of all policy measures announced since the March 2023 Budget. The Chancellor has chosen to use almost all of the £17.7 billion-per-year average improvement in the pre-measures forecast (see Chapter 4) on Autumn Statement measures. These measures raise borrowing by an average of £19.3 billion (0.6 per cent of GDP) a year from 2024-25 onwards.

Chart 3.1: Size of fiscal policy packages, 2010 to 2023

3.3 This makes it the third largest discretionary fiscal loosening on the Treasury’s scorecard since 2010. It is exceeded only by the very large support packages announced in response to Covid in the spring and summer of 2020, and by the March 2020 Budget which was delivered around two weeks before the first UK lockdown was announced. This follows the £15.6 billion a year average cost of measures the Chancellor included in the March 2023 Budget, which was the fourth costliest scorecard package during the same period.^[22]

3.4 The main policies announced in this Autumn Statement (Table 3.1) include:

• A significant reduction in the rates of National Insurance Contributions (NICs) paid by employees and the self-employed, which costs £10.7 billion (reduced to £10.0 billion once indirect effects are included) by 2028-29. The £10.0 billion estimate incorporates the revenue impact of our economic judgement that the NICs cuts increase total hours worked by 0.3 per cent, or 94,000 in full-time equivalent terms.
• A permanent 100 per cent capital allowance for qualifying business investment (known as ‘full expensing’) to replace the temporary measure announced in March 2023, which was due to end in March 2026. This lowers corporation tax receipts by £9.2 billion (increased to £10.9 billion once indirect effects are included) and raises the capital stock by 0.2 per cent by 2028-29.
• A package of welfare and other measures designed to increase labour market participation. The package increases spending by an average of £0.4 billion per year between 2024-25 and 2026-27, largely due to the DEL costs associated with the expansion of the employment schemes. It lowers spending in the final two years by an average of £0.7 billion as the WCA reforms reduce welfare spending. We estimate these measures will increase employment by around 50,000.
• A set of wider changes to public spending raise borrowing by an average of £5.8 billion between 2024-25 and 2028-29. These include an uplift in local housing allowances, aimed at addressing the increased costs of housing for benefit claimants, costing £1.7 billion by 2028-29. Other resource spending is expected to increase by an average of £4.7 billion a year, largely due to public sector pay increases and an update to the level of assumed spending beyond the Spending Review (SR) period. Near-term capital spending is also higher, in line with departmental plans, but is £4.1 billion lower in 2028-29, as the post-SR spending assumption is rolled forward.
• Other tax decisions raise revenue by £2.0 billion by 2028-29. This includes an increase in immigration fees, which raises £1.1 billion by 2028-29, an HMRC anti-avoidance and compliance package that raises an average of £1.3 billion a year, and a further one-year extension in business rates relief for the retail, hospitality and leisure sectors, which, with other measures, costs £2.5 billion next year.

3.5 The indirect effects of these policies lead to a modest boost to demand in the short-term and a small but permanent increase to potential output.^[23] The former is mainly driven by higher short-term borrowing from the cut to NICs and higher public spending, which reduces borrowing in the near term before the effect unwinds over the forecast period. The latter derives from the NICs cuts, labour supply measures and permanent full expensing. The indirect effects of policies raise borrowing in the last two years of the forecast, due to the additional debt interest costs associated with financing the measures, plus the loss in corporation tax from the extra investment induced by the full expensing measure.

Table 3.1: Total effect of Government decisions since March

Reductions in employee and self-employed NICs

3.6 The Chancellor has announced a significant reduction in the rate of NICs paid by employees and the self-employed, a tax cut that will benefit around 27 million employees and over 2 million self-employed individuals. Specifically, he has:^[24]

• Cut the rate of Class 1 NICs by 2 percentage points, from 12 per cent to 10 per cent, effective from 6 January 2024. It 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 27.3 million employees stand to benefit in 2024-25, with average annual gains of £304 for basic rate taxpayers, £647 for higher rate taxpayers, and £707 for those that pay the additional rate.
• Lowered the main rate of Class 4 NICs by 1 percentage point, from 9 per cent to 8 per cent, effective from 6 April 2024 and applying 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. This benefits 2.1 million self-employed individuals in 2024-25, with the average annual gains ranging from £117 for basic rate taxpayers, £322 for those on the higher rate, and £358 for additional rate taxpayers.
• Remove the requirement to pay Class 2 NICs for self-employed individuals with profits of at least £12,750. This will apply from 6 April 2024. We estimate there will be 1.9 million gainers in 2024-25 who will be an average of £186 a year better off.

3.7 These changes lower receipts by amounts rising to £10.0 billion by 2028-29, with an average five-year cost of £9.7 billion a year. Table 3.2 shows the main components of the costing:

• The static impact of the measure in 2028-29 is £11.1 billion – this is the amount it would raise if it did not induce any behavioural responses.^[25]
• A direct behavioural response to the measure is expected to be fewer tax-motivated incorporations (TMIs), which we estimate will be reduced by a cumulative 50,000 TMIs by 2028-29. This generates an additional £0.4 billion in receipts, with additional amounts of income tax and NICs outweighing the loss in corporation tax.
• Through indirect behavioural effects the measure is estimated to increase hours worked from both new and existing employees by 0.3 per cent or 94,000 in full-time equivalent terms. This comes through two channels:
• There is an increase in labour supply of 14,000 in full-time equivalent terms, as the 28,000 additional entrants to employment in 2028-29 join the labour force at lower-than-average hours. The earnings of the majority of those within this cohort will not be far above the personal allowance threshold and will have a low effective tax rate. There is therefore a relatively low per-person increase in income tax and NICs receipts of £0.1 billion in 2028-29.
• The increase in post-tax income is also likely to boost the number of hours worked by existing employees by 79,000 on a full-time equivalent basis in 2028-29. This offsets a further £0.6 billion of the cost of the measure in 2028-29. This higher per-person amount assumes that those that are incentivised to increase their hours worked will tend to earn more than those entering employment.
• These three dynamic effects, in aggregate, lower the cost of the NICs cut measure by £1.0 billion (close to 10 per cent). Our estimates of the size of behavioural responses are guided by the findings from empirical research on individuals’ responsiveness to changes in the marginal and average tax rates on employment (see Box 2.1).

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

3.8 This measure offsets just under a quarter of the post-pandemic personal tax rises that were announced between March 2021 and November 2022, which we now estimate will raise a combined £44.6 billion in 2028-29 (by way of comparison, this is broadly equivalent, in revenue terms, to a 10 percentage point increase in the main rate of Class 1 NICs).^[26] This is almost 50 per cent higher than our estimate in March and is driven by higher earnings growth bringing more individuals into the scope of the measures, which include the multi-year freezes in the income tax personal allowance, the income tax higher-rate threshold, and both employer and employee NICs thresholds (see Box 3.1).

Capital allowances: making 100 per cent full expensing permanent

3.9 The Government has announced that the temporary three-year 100 per cent capital allowance (known as ‘full expensing’) that it announced in March 2023 (and was due to end in March 2026) will now be in place on a permanent basis. It is forecast to lower annual receipts by £10.9 billion by 2028-29 (Table 3.3).

3.10 Full expensing (FE) means that qualifying investment in new plant and machinery that classifies as a ‘main rate’ asset can be entirely written off against taxable profit in the year that the cost is incurred.^[27] For expenditure on ‘special rate’ assets, the first-year allowance is 50 per cent. This compares to the 18 per cent and 6 per cent ‘writing-down allowance’ that would have applied from April 2026, absent this measure.^[28] There is no cap on the amount of investment that can benefit from full expensing.

Table 3.3: Costing of capital allowances: making full expensing permanent

3.11 The main uncertainty with this costing relates to how it affects the path of business investment. We assumed in March that the introduction of temporary FE would incentivise businesses to bring forward planned investment into the three-year policy window, but that, as a temporary measure, it would not impact the long-run level of the capital stock, with the brought-forward investment unwinding once the measure expired.

3.12 The costing of permanent FE is set against this existing, pre-measure, forecast for business investment (Chart 3.2). Relative to that baseline, we expect permanent FE to lower investment in the short-term, since the incentive to time-shift investment has now been removed. But it raises investment in the medium term and beyond, partly through reversing the previously assumed unwinding. Since it is a permanent reduction in the cost of capital, this measure permanently raises the capital stock by 0.2 per cent by 2028-29 (see Box 2.1).

3.13 While the impact of permanent FE on business investment is an important element of the costing, the vast majority of the cost over the forecast period relates to the static or pre-behavioural costing (Table 3.3). Of the £10.9 billion cost in 2028-29, £9.2 billion derives from the static costing, i.e. the impact of the more generous tax treatment on pre-planned investment. The dynamic impact of the additional investment that is induced as a result of the measure adds a further £1.8 billion to the cost.

Chart 3.2: Business investment as a share of GDP

3.14 Chart 3.3 shows that we are likely to have previously over-estimated the extent of the dynamic behavioural response to both March 2021’s super-deduction measure – a temporary, but more generous, two-year capital allowance that enabled 130 per cent of qualifying expenditure to be expensed – and the three-year temporary FE measure from March. Our current view is that the actual cost of both measures is likely to be lower than our initial estimates. As we explain in Box 2.4, this is partly explained by downward revisions to our view of how much additional business investment was generated by the measures, though there are other factors too. Specifically:

• We have successively lowered the costing of the super-deduction within its two-year window from an average of £12.5 billion a year originally, to £9.9 billion a year in March 2022, £8.5 billion a year in March 2023 and £6.2 billion a year now, with the latter benefiting from initial outturn data.
• Similarly, we have revised downwards the average annual costing of the three-year temporary FE measure from £9.1 billion in March to £8.3 billion now.

Chart 3.3: Super-deduction and temporary full expensing: successive costings

3.15 The effect of permanent FE and the resulting rise in business investment will continue beyond our forecast horizon as the capital stock continues to rise until it is consistent with the lower post-tax cost of capital. As a share of GDP, Chart 3.3 shows that the real terms cost of permanent FE declines from close to 0.35 per cent of GDP at the outset, to around 0.15 per cent of GDP after a decade and approximately 0.12 per cent of GDP in the long run. In today’s terms, the long-run cost amounts to just over £3 billion a year.

Chart 3.4: The long-term cost of permanent full expensing

3.16 This very large fall in costs reflects the fact that higher corporation tax deductions up front with FE mean lower deductions later. Initially, the cost of FE relative to writing-down allowances (WDAs) is high, but the relatively low 18 per cent WDA means that a pool of capital allowance claims steadily accumulates. Not having to pay WDAs each year is a saving for the Government, which we offset against the annual cost of FE to arrive at the net cost. As the pool grows, the difference between the annual FE cost and the annual WDA saving narrows, until a steady state cost is reached.

Labour supply measures

3.17 Over the next five years, the government plans to spend a total of £3.2 billion on five measures that we expect to increase employment by around 50,000 (this is on top of the changes in employment we expect from the cut in NICs). The reductions in welfare spending that result from the changes to work capability assessments mean that the package lowers borrowing in the last two years of the forecast.

Work capability assessments

3.18 Most universal credit (UC) and employment and support allowance (ESA) claimants that report a restricted ability to work due to a health condition are required to undertake a work capability assessment (WCA). The WCA is a medical process that is designed to test an individual’s capability to work and to determine if they meet the eligibility requirements for certain health-related benefits. Claimants are assessed against a set of criteria, referred to as ‘descriptors’, which cover various functional activities such as mobility, communication and daily living tasks.

3.19 Based on their assessment against these descriptors, a claimant can be considered as having: i) ‘capability to work’ (CTW), ii) ‘limited capability for work’ (LCW), or iii) ‘limited capability for work and work-related activity’ (LCWRA). ESA has three equivalent groups: ‘fit for work’ (FFW); ‘work-related activity group’ (WRAG); and ‘support group’ (SG).

3.20 In this Autumn Statement, the Government has made three changes to the WCA descriptors, to apply from September 2025 for new entrants only, and a separate change to the process around reassessments. These changes are:

• Removal of the ‘mobilising’ descriptor that enables entry into the LCWRA caseload.
• Amendment of the ‘substantial risk’ descriptor that enables entry into the LCWRA caseload.
• Amendment of the ‘getting about’ descriptor that enables entry into the LCW caseload.
• Additionally, the Government has decided that future DWP-led reassessments will take place every twelve months for new LCWRA substantial-risk claimants. If these reassessments lead to greater numbers being moved out of LCWRA, then a larger proportion of new claimants will be claiming purely on a short-term basis.^[29]

3.21 The fiscal savings arising from this policy, which amount to an average of £1.0 billion a year between 2026-27 and 2028-29, come from lower spending on the health element of UC and its equivalent in ESA. Individuals in the LCWRA group currently receive an additional £390 a month in benefits.

3.22 Removing the mobilising descriptor, amending the substantial risk descriptor, and future DWP-led reassessments will result in fewer individuals being placed in either the LCWRA group or the SG, with more being classified as within the LCW and WRAG groups. This will reduce the number of claimants receiving the health element.

3.23 Table 3.4 shows that the measure is estimated to reduce the caseload of those with the most severe incapacities by 371,000 (13 per cent) by 2028-29, with a corresponding increase of 342,000 (78 per cent) in the less severe incapacity caseload, resulting in a net reduction in the overall incapacity caseload of 29,000.^[30] No claimants drop off UC as a result of the measure, with affected claimants moving from one UC conditionality group to another.

Table 3.4: Changes to the incapacity caseload due to the WCA reform

3.24 The key uncertainty in this costing is the behavioural response:

• Given the financial gain from being assigned as LCWRA, as well as the lower level of conditionality, we expect some individuals will change their behaviour in the WCA to increase their chances of being found eligible for LCWRA against the remaining descriptors.
• We expect the policy to increase the volume of mandatory reconsiderations and appeals from those claimants that are moved out of the LCWRA caseload, and that some of these appeals will reverse that initial assessment.
• Some affected claimants will look elsewhere in the benefits system to replace the income they have lost. We assume a small proportion of those affected by the removal of the mobility descriptor will make a claim for personal independence payment (PIP) when they would not have otherwise done so.^[31]

3.25 We expect the WCA reform to raise employment by around 10,000 by 2028-29, as the loss of income from the health element (£390 a month) and higher conditionality requirements in LCW and intensive work search (IWS) incentivises these individuals to seek employment.^[32]

Other measures that affect labour supply

3.26 The remaining four labour supply measures are extensions of policies we have costed in our March 2023 Economic and fiscal outlook. Taking each in turn:

• The expansion of the Universal Support scheme increases funding for placements of disabled individuals in existing vacancies and for a ‘place and train’ programme to support them. We expect this to increase employment by around 15,000 by 2028-29 at a cumulative cost of £0.8 billion, which is partially offset by reduced welfare spending of £0.2 billion.^[33]
• The expansion of Individual Placement and Support to those with severe mental illness is expected to boost employment by around 10,000 in 2028-29.^[34] The programme entails intensive, individually tailored employment support, at a cumulative cost of £0.2 billion, which is almost fully offset by reduced welfare spending of £0.2 billion.
• Expanding access to Talking therapies is estimated to increase employment by around 10,000 by 2028-29, at a cumulative cost of £0.7 billion, which is partially offset by reduced welfare spending of £0.1 billion.^[35] The programme supports those with mild to moderate mental health conditions to enter the workforce.
• Over the next two years, the Restart scheme expansion will bring forward additional employment support to claimants in the intensive work search category who have been unemployed for 6 months, rather than 9. The cumulative cost of the expansion is expected to be £1.2 billion, which is partially offset by reduced welfare spending of £0.1 billion. We assume it will increase the number of long-term unemployed entering employment over those two years.^[36] In 2028-29, employment is therefore around 5,000 higher than would otherwise be the case.

Pillar 2: corporate taxation of global multinational profits

3.27 The UK’s implementation of the ‘Pillar 2’ corporate tax reforms, first announced at Autumn Statement 2022 and extended in this Autumn Statement, is expected to generate £2.8 billion of additional tax revenue in 2028-29.

3.28 Pillar 2 is part of an initiative led by the OECD-G20 Inclusive Framework to limit the opportunities for multinational enterprises (MNEs) to benefit from cross-border profit shifting and tax planning. It requires MNEs, with annual global revenues that exceed €750 million, to ‘top-up’ to a minimum 15 per cent effective tax rate in every country that they operate.^[37]
The intention is that this reduces the incentive for jurisdictions to engage in harmful tax competition while also reducing the attractiveness, for MNEs, of operating in very low-tax territories. Pillar 1 of the framework seeks to co-ordinate a reallocation of taxing rights with respect to a share of the profits of the largest and most profitable MNEs. That reallocation is intended to align those taxing rights more closely with the markets in which the customers of those enterprises are based. Our forecast does not include the impact of Pillar 1, which we continue to treat as a policy ambition.

3.29 Two elements of Pillar 2 were scored in November 2022 while the third and final element has been scored in this Autumn Statement. The three elements are:

• An income inclusion rule (IIR), which will be operated in the UK through a new multinational top-up tax (MTT), applies when an MNE has UK companies as well as others that also operate in a jurisdiction that has an effective tax rate below the minimum 15 per cent threshold. In such circumstances, the MNE will be required to pay a top-up to HMRC until a 15 per cent tax rate is achieved. This element accounts for £0.6 billion of the yield expected in 2028-29.
• A new domestic top-up tax (a domestic minimum tax) applies to the profits of subsidiaries located in the UK (irrespective of whether the MNE is headquartered in the UK or elsewhere). The top-up is charged when the group’s UK profits are taxed at a less than 15 per cent effective rate. This contributes £0.3 billion of the overall yield in 2028-29.
• An undertaxed profits rule (UTPR), also operated through the MTT, acts as a backstop to the IIR, taxing MNE profits that arise in countries with effective tax rates below 15 per cent that have not adopted Pillar 2 and that are also not subject to another countries’ IIR. Unlike the first two elements, the top-up tax due under the UTPR is apportioned amongst implementing countries based on each’s relative share of the MNE’s employees and assets. The UTPR costing rises to £0.5 billion in 2028-29.
• The remaining £1.3 billion of Pillar 2 yield comes from additional corporation tax due on the higher level of profits that are expected to be booked in the UK as a result of the measure, mainly due to the MTT.

3.30 Every aspect of this costing is subject to high levels of uncertainty, with some outside the usual scope of policy costings scrutiny. Specifically:

• We assign data uncertainty a ‘very high’ rating. The costing relies on data from the OECD’s country-by-country reporting (CBCR) that requires MNEs to file an annual report setting out detailed information of each jurisdiction within which they operate. CBCR information is submitted under internationally established standards, which are different to the measure required for estimating Pillar 2 profits. A second source of data uncertainty is the use of 2018-19 CBCR returns, which requires projecting ahead several years until the policy becomes effective. The uncertainty associated with this time lag is amplified by policy changes that took place in the intervening period, such as the US Tax Cuts and Jobs Act (2017) that prompted a significant repatriation of foreign profits to the US – something not reflected in the CBCR data and therefore estimated by assumption.
• The costing is based on a multiple-step model that relies on several highly uncertain assumptions that lead us to also assign a ‘very high’ uncertainty rating to this aspect of the costing. Pillar 2 rules are both novel and complicated, which makes it difficult to model its effects using the existing data. For example, several modelling adjustments are required to generate an estimate of Pillar 2 profits from the underlying data.
• There is also significant uncertainty over the behavioural response, which we also deem to have a ‘very high’ level of uncertainty. Pillar 2 is designed to prevent MNEs from minimising their global tax liability by shifting profits to lower-taxed jurisdictions, so is aimed at groups that have already demonstrated a willingness to respond to tax motivation. Precisely estimating how they might respond to this measure, and the extent of the profit-shifting impact, is therefore uncertain. Pillar 2 is also aimed at lowering the incentives for different jurisdictions to engage in harmful tax competition. While the costing does not actively include an adjustment for this, it is another source of uncertainty.
• Lastly, there is uncertainty over several aspects of the implementation and operation of Pillar 2. These include the ongoing effectiveness of the mechanisms for collecting and processing of new information, some of which will need to be obtained from overseas, the associated levels of compliance activity and the ongoing uncertainties around which countries will sign up for Pillar 2 and to which parts.

Local housing allowance

3.31 The increase in local housing allowance (LHA) rates, to equal the 30^th percentile of an area’s market rents in 2024-25, raises spending on the housing element of UC and housing benefit (HB) for eligible private renters. The LHA is a form of benefit provided to eligible individuals or families who rent their homes from private landlords. LHA rates determine the maximum amounts of UC or HB that individuals or families can claim and are based on local housing market conditions in specific areas.

3.32 LHA rates were set at the 30^th percentile in April 2020 but have been frozen since, thereby falling below the 30^th
percentile as rental prices have risen since then. This measure restores the 30^th
percentile level, bringing more rental properties into scope and increasing the maximum payable amount. This increases welfare spending by amounts that rise to £1.7 billion in 2028-29. The measure also freezes LHA rates from 2025-26 onwards, thereby eroding its generosity over time as rental prices rise.

Measures with highly uncertain costings

3.33 We assign an uncertainty rating to all certified policy costings.^[38]
The measures that we have given a ‘high’ or ‘very high’ uncertainty ratings are set out in Table 3.5. Those measures that are not discussed above are:

• From January 2024, the Government will be implementing a measure that requires certain digital platforms to report to HMRC information about the income of sellers of goods and services on their platform, in line with rules developed at the OECD.^[39] This measure is aimed at those individuals that are currently failing to declare incomes generated through these platforms. Several aspects of the costing are uncertain, including the size of the tax base and the existing level of compliance. The majority of the yield from this measure comes from assuming that the sharing of information with HMRC will prompt some individuals, that would not have done so otherwise, to meet their tax obligations. The size of this deterrent effect is highly uncertain.
• DWP fraud and error: third party data gathering compels third parties such as banks to share information with DWP that would signal potential benefit fraud. The measure generates savings by identifying fraud and error overpayments for recovery and preventing future overpayments from occurring. The data used in the costing is limited, relying on an information exchange with banks, which makes estimating the value of overpayments in scope of the measure and the impact on detection rates uncertain. This also affects modelling uncertainty, which consequently relies on assumptions, while there is also uncertainty around how those affected respond to the measure.
• From 22 November, tobacco duty on hand rolled tobacco (HRT) will increase by 10 percentage points. This is a one-off increase that is in addition to the current policy of increasing rates in line with RPI plus 2 percentage points. The main uncertainty in the costing relates to the behavioural response, which includes the fall in HRT consumption, as well as switching to related products such as cigarettes and electronic cigarettes, and the incentive to turn to the illicit market.

Table 3.5: Costings of measures with high degrees of uncertainty

Policy measures not on the Treasury scorecard

3.34 Our forecasts include the effects of several policy decisions, or related impacts that the Treasury has chosen not to present on its scorecard. These are described below, while the combined fiscal impact is shown in Table 3.1.^[40]

Home Office: visa fees and immigration health surcharge increases

3.35 The Government has announced an increase to visa fees from October 2023 and to the immigration health surcharge (IHS) from January 2024. The IHS is paid by temporary migrants who apply to come to the UK for a time-limited period of more than six months, and those who are already in the UK and apply to extend their stay. Visa fees are paid by individuals with differing intents, ranging from work and study to tourism and settlement. The IHS main and discounted rates will increase by around 65 per cent (from £624 to £1,035, and £470 to £776 respectively), while most visa fees will increase by 15 to 20 per cent. The overall revenue gain is expected to be £1.1 billion in 2028-29, of which £0.7 billion comes from additional IHS yield and £0.4 billion from visa fees.

3.36 The Government has told us that all the additional revenue generated will go towards the public sector pay awards announced in July and we have therefore added an offsetting amount to our spending forecast. Home Office taxes are expected to generate £4.9 billion in receipts by 2028-29, a five-fold increase since the introduction of the IHS in 2015-16.

3.37 The remaining non-scorecard measures include:

• Public service pensions. Changes to the employer contribution rate, known as the superannuation contributions adjusted for past experience (‘SCAPE’). The Government has lowered the discount rate applied when calculating contribution rates for public service pensions from CPI plus 2.4 per cent to 1.7 per cent, following revisions to our long-term growth forecasts. This will increase employer contributions significantly from April 2024, reducing AME spending by an average of £23.6 billion a year from 2024-25 onwards (as higher contributions reduce net spending on public service pensions). The Treasury has set aside £5.6 billion per year in RDEL spending to compensate employers. Taken together with the £18.0 billion AME forecast revisions following the ongoing 2020 valuations, these changes are fiscally neutral (see Box 4.4).
• Devolution of winter fuel payments. From April 2024, the winter fuel payment benefit will be fully devolved to Scotland. This is a close-to-neutral switch for our forecast as it replaces around £0.2 billion of DWP AME with Scottish Government AME.
• Updated levels of assumed spending beyond the Spending Review (SR) period. These increase resource spending by an average of £3.3 billion a year due to a higher starting level of spending in 2024-25 and a higher inflation forecast, though this is partially offset by lower real growth. Non-scorecard capital spending is £4.0 billion lower in 2028-29 due to the extension of the cash freeze (see Chapter 4).
• Other spending changes reflect updated departmental plans for 2024-25, as set out in the Treasury’s Public Expenditure Statistical Analyses publication in July, and our assumptions regarding underspending relative to new DEL within the SR period.

Update on previous measures

3.38 We cannot review and re-cost all previous measures at each fiscal event, with the volume being too great. But we do look at those where the original (or revised) costings are under- or over-performing, and at costings that were identified as particularly uncertain. In this section we review the costings of four sets of measures costed in previous fiscal events:

• Box 3.1 discusses the continuing impact of the multiple freezes and reductions to personal tax thresholds in recent years. High levels of inflation and the impact of fiscal drag means they are now expected to raise a combined £44.6 billion in 2028-29. By way of comparison, that is around two-thirds of the entire cost of the furlough scheme. The March 2021 Budget, where several of these measures were first announced, is already the second largest fiscal tightening since 2010, based on the estimated cost of the package at the time (see Chart 3.1). The recostings in Box 3.1 suggest it was even more of a fiscal tightening than measured at the time.
• We also provide an update on two costings from the March 2023 Budget, namely the 30-hours free childcare measure and the changes to pensions tax allowances, plus the July 2023 announcement on reducing the UK Emissions Trading Scheme cap.

Box 3.1: Fiscal implications of personal tax threshold freezes and reductions

Over the 4 fiscal events between March 2021 and November 2022, virtually 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 legislation:

• In March 2021, the income tax personal allowance (PA) was frozen at £12,750 and the higher rate threshold (HRT) at £50,270, from 2022-23 to 2025-26. The freezes were extended for a further two years in November 2022, when the upper earnings limit and upper profits limit for NICs were also frozen, meaning all these allowances and thresholds were to be frozen to 2027-28.
• In March 2022, the primary threshold and lower profits limit for NICs were increased to align with the PA (with an equivalent rise in Class 2 NICs) and then frozen until 2025-26, a freeze which was extended in November 2022 to 2027-28. In addition, the employer NICs secondary threshold was also frozen from 2023-24 to 2027-28.
• In November 2022, the additional rate threshold (ART) was lowered from £150,000 to £125,140 from April 2023 to align with the PA taper. This is the first change to the ART since it was introduced at £150,000 in April 2010, and it will remain frozen in cash terms.

Table A: Latest costings of personal tax threshold measures

Freezing thresholds, rather than raising them in line with inflation, increases tax receipts as rising wages tip ever greater numbers of workers into the tax system or onto higher rates, a trend known as ‘fiscal drag’. From one forecast to the next, higher-than-expected earnings both increase the numbers expected to be paying a higher rate of tax and the amount of tax that they pay.

Between 2022-23 and 2028-29, this set of threshold freezes means nearly 4 million additional individuals will be expected to pay income tax, 3 million more will have moved to the higher rate, and 400,000 more onto the additional rate (Chart A). This represents an increase in the number of taxpayers in each band of income tax – 11 per cent for the basic rate band, 68 per cent for the higher rate and 49 per cent for the additional rate. Relative to our March forecast, this is a respective increase in 2027-28 of 830,000, 900,000, and 43,000.

Relative to raising thresholds by CPI, this set of freezes are now estimated to raise £42.9 billion (Table A) by 2027-28 and £44.6 billion by the end of the forecast (1.4 per cent of GDP). The 2027-28 figure is £13.6 billion higher than in our March forecast, as a result of the higher and more persistent inflation in this forecast. The reduction in the employee rate of NICs in this Autumn Statement will reduce the impact of the primary threshold freeze by only around £180 million.^a Frozen thresholds are the largest contributor to the rising overall economy-wide tax burden – responsible for almost a third the 4.5 per cent of GDP increase in taxes from 2019-20 to 2028-29.

Chart A: Effect of threshold freezes on additional taxpayers and tax receipts

^a The £180 million figure is just the interaction with November 2022’s freeze in the primary threshold. It does not account for the raising of the primary threshold in March 2022. Taken together, the two changes to the primary threshold in 2022 were a net Exchequer cost. The NICs cuts announced in this Autumn Statement slightly reduces the net cost of the 2022 measures.

Lowering 30 free hours of childcare to children from 9-months-old

3.39 In March 2023, Treasury restrictions on sharing Budget-sensitive information between departments meant we were unable to cost the welfare implications of the ’30 hours free childcare’ measure using the appropriate DWP and HMRC models. Instead, we included a broad-brush estimate of our own. At this event, DWP and HMRC have re-estimated the savings based on full model runs, which has resulted in lowering the overall cost of the childcare measure by an average of £0.6 billion a year between 2024-25 and 2027-28, relative to our March estimate. The re-costings show that:

• Most of this difference (an average of £0.4 billion a year between 2024-25 and 2027-28) stems from the AME savings that are generated by those parents that switch from HMRC’s ‘tax-free childcare’ (TFC) scheme to the new, more generous scheme. By September 2025, working parents of children aged between 9 and 36 months can switch to 15 to 30 hours of free childcare that is entirely subsidised by government, in contrast to the one-fifth TFC subsidy. Our March estimate did not allow for substitution from TFC.
• A further £0.2 billion of AME savings arise from parents similarly switching from UC’s childcare element, which is higher than we estimated in March. There is a small additional saving from some UC claimants moving into work.

3.40 The reason that parents switching from a less to a more generous government scheme reduces the overall cost, is that the DEL funding for the ’30 hours’ childcare measure is unchanged from Budget 2023, whereas the higher AME savings from lower TFC and UC expenditure is captured for the first time in this forecast.

3.41 The substantial revision to this costing illustrates the importance of sharing information with those departments that have the expert knowledge required to best model the fiscal impacts.

Pensions tax allowances

3.42 In March 2023 the Government announced a package of measures relaxing the amount of pensions savings that individuals can accumulate before they incur tax charges. The two main changes were the abolition of the lifetime allowance (LTA) and an increase in the annual allowance (AA) from £40,000 to £60,000, effective from April 2023. In 2027-28, our latest estimate is that it will cost £1.6 billion, which is 43 per cent higher than our initial £1.2 billion costing. There are two explanations for the revision:

• HMRC has improved its costing model methodology so that the assumed lag between when charges are accrued and when they are eventually paid, align more closely to the tax treatment. This does not change the accrued liability but, by shortening the time lag, it increases the amount of receipts foregone within the forecast period, as a result of the measures. This element contributes £0.3 billion (58 per cent) of the combined difference between the original and revised costings.
• Outturn data show an uptick in LTA and AA charges, relative to what we assumed in March. LTA charges in 2021-22 rose to £497 million, 27 per cent higher than in 2020-21 and 14 per cent higher than our March forecast (Chart 3.5). In 2021-22, schemes reported 11,660 LTA charges through ‘accounting for tax’ (AfT), a third higher than 2020-21. Similarly, the number of AA charges reported through AfT increased from 17,280 in 2020-21 to 49,380 in 2021-22, while the number of self-assessed individuals reporting pensions contributions that exceeded the AA also rose, from 43,870 to 53,330 over the same period.^[41] This component, combined with other new data, accounts for £0.2 billion (42 per cent) of the overall difference.

Chart 3.5: Lifetime allowance charges: outturn and forecast

Reduction to the UK Emissions Trading Scheme cap

3.43 In July 2023, the Government announced the pathway for a net-zero consistent cap to allowances for the UK Emissions Trading Scheme (ETS). This resulted in a 60 per cent reduction in allowances in the final year of the forecast.^[42]
While this is a new policy announcement since our March forecast, we have incorporated its effects into our pre-measures forecast rather than through the Treasury’s scorecard.

3.44 This is because to estimate the cost of the measure requires assessing its impact on the price of allowances. However, the impact of the measure on the price of allowances will already be captured in our pre-measures forecast of ETS revenues. The carbon price moved previously in response to the series of consultations and statements on the likely direction of Government policy between 2021 and early 2023.^[43]
It is probable therefore that much of the impact of this measure was actually ‘priced-in’ at the time of our March forecast, even ahead of the final announcement in July. To calculate the cost of the measure would require us to now generate a counterfactual price series in the absence the measure, which would be highly uncertain. This has no impact on our overall receipts forecast which uses the new net-zero consistent path for allowances and the latest carbon futures forecast for prices.

Policy risks

3.45 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.

3.46 In addition to the full expensing policy described above, risks that have crystallised and are now reflected in our forecast include:

• Home Responsibilities Protection (HRP) correction. In September 2023, HMRC began sending letters to invite affected customers to claim the arrears they are owed due to historic underpayments of the state pension. The exact number of claimants affected, their average arrears, and take-up rates remain uncertain, but this is now reflected in our welfare forecast, costing a total £1.9 billion over the forecast period.
• The UK-Australia and UK-New Zealand trade deals, signed on 16 December 2021 and 28 February 2022 respectively, removed tariffs on all UK goods exported to Australia and New Zealand, and nearly all Australian and New Zealand exports to the UK, subject to meeting ‘rules of origin’ requirements. These came into force on 31 May 2023 and are expected to lower customs duties by around £45 million a year over the forecast. The Government’s analysis suggests the two deals might increase the level of real GDP by a combined 0.1 per cent by the 2035.^[44]

3.47 Risks that have evolved or that are new since March include:

• A tax on vaping. On 12 October, the Government launched a consultation considering various measures to tackle smoking and vaping, including "exploring a new duty on vapes", with the aim to implement these as soon as possible after the consultation closes on 6 December.^[45] We will incorporate the impacts of these measures into our forecast once they are sufficiently certain. The costing is likely to be subject to several areas of uncertainty.
• 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 could lead to “an increase in UK GDP of £1.8 billion in the long run”.^[46] If we take the long run to be 15 years after joining, this equates to around 0.04 per cent of GDP.
• On 30 March, the Government published an exploratory consultation considering "a range of potential policy measures to mitigate carbon leakage risk in the future and ensure UK industry has the optimal policy environment to decarbonise".^[47] Potential policies included a carbon border adjustment mechanism, which would introduce a carbon price on imported products reflecting both the carbon emitted in their production and the gap between the carbon price in their country of origin and the price had it been produced in the UK. As yet, there is insufficient detail to incorporate this into our forecast. It is another costing that is likely to have several highly uncertain components.

Chapter 4: Fiscal outlook

Introduction

4.1 This chapter:

• notes classification issues affecting our forecast (from paragraph 4.5);
• describes the outlook for public sector receipts (from paragraph 4.6) and public sector expenditure (from paragraph 4.40);
• 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 4.75); and
• summarises key uncertainties and risks to the fiscal outlook (paragraph 4.85).

4.2 The forecasts in this chapter start from the estimates of 2022-23 outturn data published by the Office for National Statistics (ONS) on 20 October. We then present an in-year estimate for 2023-24 that makes use of ONS outturn data for April 2023 to September 2023 (but not the October outturn data released on 21 November). Finally, we present forecasts for 2024-25 to 2028-29.^[48]

4.3 Throughout this chapter we compare our latest forecast with those from our most recent Economic and fiscal outlook (EFO) published in March 2023. Since March, ONS revisions have resulted in a 1.9 per cent increase to the level of GDP in 2022-23. As this level shift has no impact on the cash level of receipts and spending, but does affect fiscal aggregates expressed as a per cent of GDP, we have, where useful to explain changes, restated our March forecast for this change.^[49]

4.4 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 then presents scenarios in relation to some of these key sources of uncertainty.

Classification and other statistical changes

4.5 No ONS classification decisions since our March forecast have affected this forecast. The ONS has now classified the component parts of Bulb following the transfer of its customers to Octopus Energy. However, this has not yet been incorporated into ONS statistics and so we have not included the consequences of future Bulb flows in this forecast. We expect these flows to have very little effect on the forecast.

Public sector receipts

Summary of the receipts forecast

4.6 In 2022-23, total public sector receipts as a share of GDP reached 40.1 per cent, a 3.2 percentage point increase on their pre-pandemic level of 36.8 per cent of GDP in 2019-20. Receipts are forecast to continue to grow faster than nominal GDP, rising as a share of the economy by a further 1.5 percentage points to reach 41.6 per cent of GDP by end of the forecast in 2028-29. Even after the significant reductions in personal and corporate taxes announced in this Autumn Statement, the tax burden (the ratio of National Accounts taxes^[50]
to GDP) is forecast to reach a post-war high of 37.7 per cent of GDP in 2028-29, 4.5 percentage points above its pre-pandemic level of 33.1 per cent of GDP in 2019-20 (Chart 4.1).

4.7 Relative to our March forecast, the tax burden has been revised down by 0.7 percentage points this year and by 0.3 percentage points in 2027-28. This is due to largely offsetting effects from historical revisions, forecast changes and policy measures:

• The level of nominal GDP has been revised up by the ONS by 1.9 per cent in 2022-23, which, before any other changes, would have reduced our March forecast of the tax-to-GDP ratio by 0.7 percentage points every year across the forecast period. This ‘restated March’ forecast using the higher GDP outturn is shown in the solid yellow line in Chart 4.1.
• Receipts in 2022-23 and in the first half of 2023-24 have been much stronger than expected in March, largely driven by higher inflation and a larger nominal economy (see Box 4.1). This has led us to revise the 2023-24 tax-to-GDP ratio up by 0.1 percentage points relative to the restated March forecast.
• The composition of nominal GDP growth over the forecast is also now expected to be more tax rich, with stronger nominal wage growth, nominal consumption and profits. As a result, in our pre-measures forecast, the tax burden is expected to rise sharply in 2024-25 driven mainly by income tax (as average nominal earnings rise by 3.3 per cent while income tax thresholds remain frozen) and corporation tax (as the full-year impact of the higher tax rate introduced in April 2023 takes effect). This leaves the tax burden, before policy announcements, 1.1 percentage points above the restated March forecast in 2027-28.
• The impact of policies announced in this Autumn Statement (including their indirect effects), namely the reduction in the rates of National Insurance contributions (NICs) paid by employees and the self-employed, and making full expensing permanent (partially offset by tax increases from visa fees, the immigration health surcharge and the introduction of the new Pillar 2 tax), reduces the tax burden on average by 0.5 percentage points from 2024-25, to leave it 0.4 percentage points higher than our restated March forecast in 2027-28.

Chart 4.1: National Accounts taxes as a share of GDP

4.8 The tax-to-GDP ratio at the forecast horizon in 2028-29 is expected to be 4.5 percentage points higher than it was in 2019-20 (Chart 4.2). This reflects both pre-existing trends and subsequent changes:

• The tax system inherited by Chancellor Sunak in March 2020 would have increased the tax-to-GDP ratio by 0.2 percentage points, due largely to fiscal drag.^[51]
• Underlying forecast changes since then raise the ratio by 2.5 percentage points over the nine-year period, largely reflecting a more tax-rich composition of economic activity.
• Policies announced between March 2020 and 2022, including maintaining the rate of corporation tax at 19 per cent and then raising it to 25 per cent, introducing the health and social care levy, and freezing income tax thresholds until 2025-26, would have added 2.1 percentage points to the tax burden in 2028-29.
• The net impact of the November 2022 Autumn Statement measures increases the tax-to-GDP ratio by 0.3 percentage points in 2028-29. This includes tax rises announced by Chancellor Hunt (mainly the extension of the income tax and NICs threshold freezes until 2027-28), which more than offset the tax cuts retained from Chancellor Kwarteng’s Growth Plan, notably the abolition of the health and social care levy.
• Measures announced in the March 2023 Budget reduce the tax-to-GDP ratio by 0.1 percentage points in 2028-29, mainly from the fuel duty freeze and more generous pensions tax allowances.
• Measures announced in this Autumn Statement decrease the tax burden by 0.6 percentage points in 2028-29.

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

Change in receipts since our March 2023 forecast

4.9 Relative to our March forecast, and including the impact of Autumn Statement measures, receipts are up by £40.8 billion (3.9 per cent) in 2023-24, and then up by an average of £50.7 billion a year (4.4 per cent) between 2024-25 and 2027-28. This is driven by stronger-than-expected outturn this year, and a stronger forecast for the nominal drivers of the key tax bases, both driven, mainly, by higher inflation (further explored in Box 4.1 below). This has been partially offset by the impact of policy measures announced in this Autumn Statement, and by lower energy, carbon and equity prices (which reduce oil and gas revenues, emissions trading scheme revenues, and capital gains tax receipts, respectively).

Table 4.1: Receipts: changes since March 2023

Box 4.1: Why are receipts stronger since March?

In our pre-measures forecast, tax receipts have increased in all years relative to our March forecast, and by £57.0 billion in 2027-28 (this excludes fiscally-neutral VAT refunds and environmental levies). This improvement is spread across the main taxes with increases in income tax and national insurance contributions, onshore corporation tax, and VAT, offset by falls in smaller taxes (Chart A).

Chart A: National accounts taxes (excluding PSNB neutral): Changes since March

A large portion of the upside revision occurs this year, reflecting strength already seen in the outturn statistics. In the first half of this year, national accounts taxes, excluding fiscally neutral items, are £15.1 billion above our profile from the March forecast. This is largely due to the impact of higher-than-expected inflation on the main taxes but also reflects some more technical factors. In the first half of this year:

• Income tax and NICs are up £6.0 billion compared to our March profile. Stronger nominal earnings growth than anticipated in March, along with one-off payments and backdated pay awards in the NHS and the Civil Service, has boosted these receipts. In addition, strong growth in pension income (reflecting inflation or earnings-related rises) and greater use of pension flexibility have also boosted receipts.
• Corporation tax is up £7.0 billion compared to our March profile. In addition to higher-than-expected non-financial and financial sector profits, this also reflects what now looks like a pessimistic assumption in our March forecast that strength in 2022-23 would not be maintained. In addition, the sectors that are generally the bigger taxpayers (financial, retail and professional services) continue to perform well.
• VAT is up by £6.4 billion compared to our March profile. On top of higher inflation, this also reflects our overestimation of the negative impact of the cost-of-living crisis on VAT receipts, as both the level of consumption and the share of standard rated items have been higher-than-expected.
• All other taxes combined are down by £4.3 billion compared to the March profile, driven by offshore corporation tax, energy profits levy and electricity generators levy which are down by £1.0 billion, £1.1 billion and £0.9 billion, respectively, due to lower energy prices.

For the second half of this year our forecast implies continued upside strength, with taxes £19.5 billion higher than expected in our March forecast. Again, the continued impact of high inflation on the main tax bases is expected to be the key contributor to this strength. In the second half of the year:

• Income tax and NICs are expected to be £9.5 billion higher than our March profile. While earnings growth is expected to slow in the second half of the financial year, it remains stronger than we anticipated in March. Stronger bonus growth than assumed in March is also expected to boost receipts in the final months of the financial year.
• Corporation tax is forecast to be £9.7 billion higher than our March profile. While there may be some effect from slowing profit growth in the remainder of the year, recent cash receipts outturns have so far shown little sign of moderating. We also expect some upward revisions to accrued receipts for the first half of the year.^a
• VAT receipts are forecast to be £4.6 billion above our March profile, as we expect the current strength to continue, albeit to a lesser degree as the growth of nominal consumption is forecast to slow in the second half of the year.
• All other taxes combined are forecast to be £4.3 billion below our March profile, continuing the weakness from the first half of this year.

Overall, due to these factors we forecast taxes in full-year 2023-24 will be £34.6 billion higher than in our March forecast. In addition to this strength in 2023-24, further upward revisions compared to our March forecast total an additional £22.4 billion in receipts by 2027-28. The drivers of this further strength between 2023-24 and 2027-28 are:

• IT and NICs improves by an additional £17.1 billion by 2027-28 relative to March. Nominal earnings growth is expected to be stronger through the forecast, particularly in 2024-25. Combined with income tax and NICs threshold freezes, this boosts fiscal drag significantly.
• VAT increases by an additional £6.3 billion by 2027-28. On top of a higher nominal GDP, driven by higher inflation, this reflects a shift to higher standard-rated consumption as the impacts of high inflation subside and consumption shifts away from the lower- or zero-rated necessities such as food, energy and rent.
• Other taxes add an extra £2.2 billion, as the deficit against our March forecast reduces from £8.6 billion this year to £6.4 billion in 2027-28.
• Offsetting this, onshore corporation tax surpluses relative to March fall by £3.2 billion by 2027-28, largely reflecting lower profit growth particularly in 2024.

As shown in Chart B below, across all years of our forecast, higher and more persistent inflation is by far the largest driver of these upward revisions to taxes, by increasing the nominal tax base. The average upward revision since March in our pre-measures tax forecast is £44.4 billion a year (diamonds). Of this, £28.0 billion is explained by a larger tax base (nominal GDP). This is the net result of lower real GDP growth, which compared to our March forecast reduces receipts by £13.4 billion (yellow bars) and higher inflation, which increases receipts by £41.4 billion (green bars).^b The rest is explained by a higher effective tax rate, which increases receipts by £16.4 billion (blue bars), about half of which is fiscal drag driven by strong earnings, and IT and NICs threshold freezes, which increase receipts by an average of £7.2 billion a year compared to our March forecast (explained in Box 3.1).

Chart B: National accounts taxes (excluding PSNB neutral): Changes since March

a) Accrued corporation tax for 2023-24 is based on both cash outturn and forecasts of cash receipts that will be paid in future months and accrued back. The latest published data for accrued CT has already moved away from March EFO forecasts to allow for higher cash receipts in future months. Incorporating the November EFO forecast (particularly higher receipts from small companies) could potentially raise accrued receipts further in the first half of 2023-24.

b) Using the GDP deflator as the broadest measure of inflation.

Tax-by-tax analysis

Income tax (excluding self-assessment) and NICs

4.10 PAYE income tax and NICs receipts are expected to rise by £25.5 billion (6.6 per cent) in 2023-24 to reach £414.4 billion. From 2024-25 onwards, we expect receipts to grow, on average, by 3.9 per cent a year to reach £501.3 billion (15.6 per cent of GDP) by the end of the forecast period.^[52] This reflects strength in receipts this year, nominal earnings growth of 13.5 per cent over the rest of the forecast and the ongoing freezes to income tax thresholds that generate powerful fiscal drag (see Box 3.1 for detail on the impact of these freezes). As shown in Chart 4.3 (which captures all income tax and NICs), these lead to an increase over the forecast period in the income tax and NICs effective tax rate (ETR) – even after the impact of the reduction in NICs rates announced in this Autumn Statement.

4.11 Relative to our March forecast, all non-SA income tax and NICs receipts have been revised up by £12.5 billion (3.2 per cent) in 2023-24, reflecting stronger outturns, and then by an average of £17.4 billion (4.0 per cent) a year between 2024-25 and 2027-28. As shown in Table 4.2, the main sources of these changes are:

• On a pre-measures basis, the tax base (income from wages and salaries) has been revised up by an average of 3.3 per cent each year, raising receipts by an average of £15.0 billion a year. This primarily reflects our forecast of higher and more persistent inflation and much stronger nominal earnings growth.
• The effective tax rate (ETR) (as shown in Chart 4.3), has been revised up by an average of 0.2 percentage points a year, raising receipts by an average of £7.7 billion.^[53]This largely reflects higher fiscal drag due to the stronger earnings growth in this forecast. In addition, the ETR is boosted by the distributional effect of stronger bonus growth which is concentrated among high earners.^[54]
• Autumn Statement measures reduce revenue by an average £8.0 billion in each year of the forecast, with the bulk from foregone NICs (both employee and self-assessed) receipts. The reduction in NICs rates for employees and self-employed alone costs £8.6 billion on average over the forecast period.
• Indirect effects of the labour supply packages outlined in the previous chapter add an average £1.7 billion of receipts to each year of the forecast. These measures boost labour supply through reducing inactivity and incentivizing those already in employment to increase their hours of work.

4.12 Other income tax includes the charges from the Lifetime Allowance (LTA) (abolished in April 2023) and the Annual Allowance (increased from £40,000 to £60,000 from April 2023). The costings of these policies have been revised significantly with the availability of new data. We expect the cost of these policies to be on average £0.5 billion more per year of the forecast than we had assumed in our March forecast.

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

Chart 4.3: Income tax and NICs receipts as a share of GDP, wages and salaries as a share of GDP, and the effective tax rate (ETR)

Self-assessment (SA) income tax

4.13 Self-assessed income tax receipts (which relate to income from sources including dividends, savings, property and self-employment) are forecast to grow by an average 9.2 per cent over the forecast. Most SA payments are made after the tax year has ended, so receipts largely relate to tax liabilities in the previous year. There is a particularly strong rise in tax receipts in 2024-25 of 18 per cent driven by a large rise in savings income in 2023-24, due to higher interest rates, and fiscal drag due to high self-employment income growth combining with frozen tax thresholds.

4.14 Relative to our March forecast, we have revised receipts up by an average of £2.9 billion a year.

• Forecast changes add an average £2.5 billion to each year of the forecast. This reflects changes to the tax base, including upward revisions to savings income due to higher interest rates – adding an average £1.1 billion to each year of the forecast – and self-employment income adding an average £0.3 billion to each year of the forecast.
• Autumn Statement measures add an average £0.2 billion to receipts in each year of the forecast. These measures include some small knock-on effects from the reduction in NICs rates and measures to improve the collection of tax debt that has built up.
• Indirect effects of policies add around £0.1 billion in receipts, mainly from the reduction in NICs rates incentivizing increased employment and increased hours among those already working.

VAT

4.15 VAT receipts are expected to increase by 6.9 per cent to £173.2 billion this year, broadly in line with nominal GDP which grows by 6.8 per cent. Receipts are then forecast to increase by an average of 3.8 per cent a year over the remainder of the forecast. This growth is stronger than the 3.4 per cent annual growth in nominal GDP, because the composition of consumption is expected to shift towards standard-rated goods taxed at 20 per cent. The standard-rated share of household consumption fell this year by 1.7 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 consumption to increase back to its historical levels, reaching 49.5 per cent in 2028-29.

4.16 The ratio of VAT receipts to nominal GDP has increased from 5.4 per cent in 1999-2000 to 6.4 per cent in 2022-23, as shown in Chart 4.4. This is explained by an increase in the standard rate of VAT from 17.5 per cent to 20 per cent, changes in the composition of expenditure, and a reduction in the VAT gap from 12.5 per cent in 1999-2000 to 5.4 per cent in 2021-22.

4.17 The VAT gap (a measure of compliance) is the difference between actual VAT receipts collected and the VAT theoretical liability, estimated using national accounts data. This significant fall in the VAT gap has been driven by HMRC’s program of digitalisation of the VAT system, falling use of cash in the economy, and a range of compliance measures. We forecast the VAT-to-GDP ratio to increase slightly from 6.4 per cent this year to 6.5 per cent in 2028-29, driven by increased standard-rated consumption, as described above. We assume that the VAT gap remains broadly flat in per cent terms across the forecast period as it becomes increasingly difficult to reduce further. The possibility of further falls in the VAT gap represents an upside risk to our receipts forecast.

Chart 4.4: VAT rates, gap and receipts as a share of GDP

4.18 Relative to our March 2023 forecast, receipts have been revised up by £11.1 billion (6.9 per cent) in 2023-24, reflecting the stronger-than-expected outturn so far this year, which was driven by a bigger nominal economy due to higher inflation. In future years receipts have been revised up by an average of 14.5 billion (8.2 per cent) reflecting the combination of a higher starting point and stronger growth in nominal consumption, driven by higher inflation. The direct impact of measures announced in this Autumn Statement increases VAT receipts, on average, by £0.2 billion a year.

Onshore corporation tax

4.19 Onshore corporation tax (CT) receipts have repeatedly performed more strongly than we have expected in recent years, due to a surprising resilience in underlying profits and in the average tax rate paid on those profits. We have seen further unexpected strength again so far this year, in addition to the expected effect of the rise in the main rate from 19 to 25 per cent in April 2023.

4.20 As a result, our estimate for accrued onshore CT receipts in 2023-24 has been revised up by £17.0 billion relative to our March forecast. Cash receipts were up £8.8 billion in the first half of 2023-24. We expect that much of the strength in receipts will persist over the rest of 2023-24, although it may be slightly dampened by slower forecast profit growth in the remainder of the year. The main drivers of the change in the forecast this year and over the medium term are:

• Profits from non-financial and financial sector companies have been stronger than expected in 2023. We are now assuming a rise of 8.4 per cent and 10 per cent, respectively, compared with forecasts of a 1.2 per cent fall and zero per cent change in March. This boosts our forecast of receipts relative to March by £7.5 billion in 2023-24. Profit growth from non-financial companies is expected to be weaker in 2024 than in our March forecast, which slows receipts growth next year.
• The pre-measures effective tax rate (ETR) has also been revised upwards since March, as we have now seen a sustained rise in receipts relative to measures of profits in the National Accounts, which is over and above the impact of the rise in the main rate of corporation tax. This has been broad-based with rising ETRs in sectors that have historically been big payers of CT, such as financial services, retail and professional services. We therefore assume that much of this higher ETR will persist across the forecast period.
• The direct effects of Autumn Statement measures reduce receipts by £9.9 billion in 2027-28. This mainly reflects the announcement that full expensing of capital allowances will be made permanent rather than finishing at the end of 2026-27.
• The indirect effects of Autumn Statement measures raise receipts in the short run, but lower receipts in the latter part of the forecast. Initially, making full expensing permanent removes the incentive to time-shift investment, reducing investment and raising corporation tax. In the final years of the forecast, the measure will boost investment and lower corporation tax (by £1.4 billion in 2027-28).

Table 4.3: Onshore corporation tax: changes since March 2023

4.21 As a result of these factors, onshore CT receipts are forecast to jump from 2.9 per cent of GDP last year to 3.4 per cent of GDP this year, and then rise to around 3.6 per cent of GDP in each year from 2024-25 onwards – which would be their highest level since corporation tax was introduced in 1965. Chart 4.5 shows that a significant driver of this is an increase in the effective tax rate (ETR). This increase in the ETR is driven in part by the six percentage point rise in the main rate of CT from April 2023, which is only partly offset by the more generous capital allowances regime resulting from the permanent full expensing measure. By contrast, the cuts in the main rate of corporation tax between 2010-11 and 2017-18 were largely offset by base-broadening measures restricting allowances and reliefs (for example, on interest and loss relief), keeping the overall ETR relatively stable over this period. Further, as described above, receipts over the past couple of years have been stronger than implied by profits as measured in the National Accounts across most sectors, which we assume is largely maintained across the forecast.

Chart 4.5: Onshore corporation tax rates and receipts as a share of GDP

Oil and gas receipts

4.22 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 place a 75 per cent tax on North Sea profits until the end of March 2028. We expect these taxes to raise £6.1 billion in 2023-24 – a £3.7 billion fall on receipts in 2022-23 – in large part due to lower expected energy prices, and (to a lesser extent) lower production and higher expenditure by operators, including for investment and decommissioning. We expect receipts to fall relatively steadily from 2023-24 to reach £2.1 billion by 2028-29, as energy prices and production declines, and the EPL comes to an end in March 2028.

4.23 Relative to our March forecast, we have revised receipts down by an average of £2.4 billion a year between 2023-24 and 2027-28, primarily due to the lower path of gas prices. While the market response so far has been muted, the conflict in the Middle East remains a source of uncertainty surrounding the outlook for global energy prices, and the oil and gas receipts forecast.

Fuel duties

4.24 Fuel duty is expected to raise £24.4 billion this year before rising to £28.2 billion in 2024-25, driven by the current stated policy to reverse the temporary 5p cut on the 23^rd of March next year and index the duty rate by RPI from April 2024. Receipts are then expected to rise at a slower rate as the upward effect from higher duty rates and increasing vehicle use is increasingly offset by a growing electric vehicle (EV) share.

4.25 In practice, the Government’s indexation policy has rarely been implemented. To illustrate the risk associated with not implementing this policy, Chart 4.6 shows our forecasts for fuel duty revenues, and the extent to which our latest forecast is boosted by the stated reversal of the temporary 5p cut and the resumption of RPI indexation of the duty rate. On average, these policies add £4.9 billion (0.2 per cent of GDP) a year to receipts between 2024-25 and 2028-29, and £6.2 billion in 2028-29, relative to the duty rate remaining constant in cash terms. In Chapter 5 we show how this affects headroom against the Chancellor’s fiscal targets.

Chart 4.6: Fuel duty: forecasts versus outturn

4.26 Relative to our March forecast, receipts have been revised up by an average of £2.4 billion a year, of which £1.1 billion is explained by higher inflation pushing up the indexed rate of duty. The remainder is explained by downward revisions to our efficiency assumptions for petrol and diesel vehicles, and to the EV share, both of which increase overall fuel consumption.

4.27 Previously we assumed that the efficiency improvements for petrol and diesel vehicles would continue the trend displayed between 2010 and 2020, when they increased from 34 to 38 miles per gallon (mpg) for petrol cars and from 41 to 46 mpg for diesel cars.^[55]
However, latest outturn data suggests that the improvement in efficiency stalled after 2020, and we assume it is unlikely to accelerate again as manufacturers shift their focus to developing EVs. We now use the Department for Transport’s projections for improvements in efficiency^[56]
which are consistent with this assumption, keeping both petrol and diesel car efficiency relatively flat throughout the forecast at around 38 and 45 mpg, respectively. This increases fuel duty receipts by £1.2 billion a year, on average.^[57]
The downward revision to our assumption regarding take up of EVs, explained in Box 4.2, increases fuel duty receipts by a further £0.7 billion per year, on average.

Box 4.2: Updated electric vehicle assumptions and their fiscal implications

As the share of electric vehicles (EVs) increases, receipts from fuel duty, and to a lesser extent Vehicle Excise Duty (VED), will decline. Fuel duty raised £25.1 billion (1.0 per cent of GDP) in 2022-23, has been falling as share of GDP since 1998-99, and is expected to fall to zero in cash terms as EVs replace petrol and diesel cars in line with the Government’s planned transition to net zero by 2050.^a

In previous forecasts the EV share of new car sales had repeatedly exceeded our expectations, increasing from 0.5 per cent of new car sales in 2017-18 to 13.6 per cent in 2021-22. To reflect this, in our March 2022 forecast, we made an upward revision to our assumption on the pace of EV take-up which we assumed would rise to 59.6 per cent of new car sales by 2026-27.^b However, in 2022-23, growth in EV take-up slowed, accounting for just 16.5 per cent of new car sales, which was more than one percentage points below our March 2023 forecast of 17.7 per cent. Evidence points to a number of reasons for the slowdown:

• The decline in EV price gap at point of purchase relative to internal combustion engine vehicles (ICEVs) has begun to slow, with a 15 percentage point fall in the two years to March 2022 but only a 6 percentage points fall since.^c The generally higher upfront costs of EVs relative to ICEVs will likely still be disincentivising many consumers, especially purchasers using car finance as interest rates are significantly higher than we had anticipated in 2022. In the absence of low cost EVs, the steep sales growth of the past years, boosted by (usually high-income) early adopters, is expected to slow.^d
• EVs have lower running costs for consumers who can charge vehicles at home. However, the cost advantage of EVs charged away from home is significantly less and can become negative, and the availability of public charging points seems to be a concern for many drivers.^{e f} Moreover, petrol and diesel prices have declined from the spike in 2022, due to a combination of both wholesale price falls and fuel duty cuts, though are still high relative to the past.

The main policy driver for EV uptake is now the Zero Emission Vehicle (ZEV) mandate which takes effect in January 2024. We have therefore revised our EV assumption to match the path of the mandate over the forecast horizon. The mandate sets a minimum share of cars and vans sold by each manufacturer to be zero emission. We judge sales are unlikely to materially exceed this across the forecast horizon due to flexibilities that allow trading of allowances and borrowing against future allowances in the first three years of the mandate.^g Furthermore, the Government recently announced a 5-year delay on the ban of new ICEV sales, from 2030 to 2035, which may result in some consumers delaying a switch to EVs.

The announced ZEV mandate reduces our forecast of EV uptake (compared to March 2023) from 25 per cent to 18 per cent in 2023, and from 67 per cent to 38 per cent in 2027 (Chart C). This has increased the fuel duty forecast on average by £0.7 billion a year and the VED forecast by £0.1 billion a year since March.

Chart C: Electric vehicle share of new car sales

a) OBR, Fiscal risk and sustainability, July 2022 (p. 140)

b) OBR, Economic and fiscal outlook, March 2022 (Box 3.3)

c) Auto Trader, The road to 2030, accessed 18 November 2023

d) Reuters, EV demand in Europe to slow as customers await affordable electric cars, accessed November 2023

e) World Economic Forum, Electric vehicles: The 3 main factors holding back sales, accessed November 2023

f) RAC, More drivers than ever expect to go electric next time but many likely to delay making the switch, accessed November 2023

g) Department for Transport, Zero emission vehicle (ZEV) mandate consultation: summary of responses and joint government response, October 2023

Other receipts

4.28 Stamp duty land tax receipts are expected to fall by 21.9 per cent this year to £13.0 billion and by a further 2.6 per cent to £12.7 billion the following year, reflecting the downturn in the housing market coupled with the September 2022 policy measure to increase the nil-rate thresholds that will come to an end in March 2025. The remainder of the profile is driven by housing transactions which rise steadily over the forecast, with receipts reaching £22.2 billion by the end of the forecast. Relative to our March forecast, we have revised up receipts in the near term, but down from 2025-26 reflecting lower expected housing transactions.

4.29 Capital gains tax (CGT) receipts are expected to be £16.5 billion in 2023-24, a 2.5 per cent fall from last year, and to fall by a further 4.2 per cent in 2024-25. Receipts are then expected to rise over the remainder of the forecast following the profile of equity prices. Relative to March, receipts have been revised down by an average £4.4 billion each year of the forecast due to downward revisions to the equity price forecast and a smaller proportion of self-assessment receipts (paid on last year’s liabilities) attributed to CGT. Property disposals have also come in lower than expected this year.

4.30 Inheritance tax (IHT) is expected to raise £7.6 billion this year, an increase of 7.5 per cent on last year. We then expect it to fall slightly before picking back up over the rest of the forecast. As with CGT, 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 March, receipts have been revised up across the forecast by an average of £0.4 billion, in large part due to higher receipts than anticipated this year.

4.31 Alcohol duty receipts are expected to raise £13.0 billion this year, up by £0.6 billion from last year, driven by the 10.1 per cent duty rate increase in August this year, partially offset by lower consumption.^[58]
Receipts are expected to rise throughout the forecast period, reaching £17.1 billion in 2028-29, as the duty rate increases with RPI indexation and consumption remains relatively flat. Compared to our March forecast, receipts have been revised down between 2023-24 and 2025-26, due to lower-than-expected consumption in recent outturn data (driven by an increase in duty rates and cost-of-living pressures) and revised up in the remaining years as the impact from higher duty rates (due to higher inflation) outweighs the reduction in volumes.

4.32 Tobacco duty receipts are expected to raise £8.9 billion this year, down £0.5 billion (5.2 per cent) on 2022-23. Receipts are expected to increase slightly in the next two years, as the positive impact from higher duty rates (including the one-off increases in hand rolling tobacco announced in this Autumn Statement) outweigh the downward trend in tobacco consumption, before starting to decrease as the impact from the higher duty rate is no longer sufficient to offset declining consumption. Relative to our March forecast, receipts have been revised down by £1.5 billion this year and by an average of £0.9 billion a year thereafter, reflecting lower consumption partially offset by the higher duty rates.

4.33 Vehicle excise duty (VED) is expected to raise £8.0 billion this year, up by £0.7 billion (9.1 per cent) on 2022-23, and is expected to increase throughout the forecast period, driven by an increasing tax base (car fleet), higher duty rates and the extension of VED to electric vehicles from 2025 onwards. Receipts have been revised up by an average of £0.4 billion a year since our March forecast, more than explained by higher RPI inflation increasing the duty rate. We have revised our EV assumptions as explained in Box 4.2.

4.34 The emissions trading scheme is expected to raise £6.2 billion this year, before it falls sharply in the next two years, due to lower carbon prices, to £2.9 billion in 2025-26. It steadily decreases in the final three years of the forecast driven by the decreasing number of allowances announced in the summer (discussed in more detail in Chapter 3). Our forecast assumes that the carbon price follows the futures market price in the next two years and remains constant thereafter. Compared to our March forecast, receipts have been revised down by £2.5 billion a year, on average, due to lower carbon prices and allowances.

4.35 The electricity generator levy (EGL) is expected to raise £1.4 billion in 2023-24, falling to £0.1 billion by the end of the forecast (EGL is due to finish by the end of 2027-28). The path primarily reflects the projection for wholesale electricity prices. Relative to March, receipts have been revised down by an average £1.5 billion a year between 2023-24 and 2027-28, primarily reflecting lower wholesale electricity prices.

4.36 Receipts from environmental levies are expected to yield £9.6 billion in 2023-24 and have been revised up by an average of £3.4 billion a year across the forecast. In the near term, this is largely due to lower energy prices affecting the transfers to electricity producers covered by the contracts for difference scheme. In the final years, the changes are largely due to the increase in the projected costs of the capacity market scheme due to higher-than-projected clearing prices. These levies are recorded equally in our tax and spending forecasts, so are neutral for public sector net borrowing.

4.37 VAT Refunds are expected to reach £28.0 billion this year, then grow throughout the forecast in line with government consumption, reaching £31.9 billion in 2028-29. Since March, the forecast has been revised up, on average, by £3.1 billion a year, mainly due to higher government consumption. Like environmental levies, VAT refunds are also offset in spending and therefore neutral for public sector net borrowing.

4.38 Interest and dividend receipts include income from the government’s financial assets such as student loans and bank deposits. The path of receipts in part reflects the profile of interest rates, which are forecast to average around 5 per cent both this year and next before falling. It also reflects the path of RPI-linked student loan accrued interest. Relative to March, we have revised receipts up by an average of £3.1 billion a year between 2024-25 and 2026-27 due to these two factors.

4.39 Business rates are expected to be £29.5 billion this year, rising to £38.0 billion in 2028-29. Compared to March, forecast revenue is £0.4 billion lower in 2023-24 and £2.4 billion lower in 2024-25, but is higher by around £1.3 billion a year thereafter. A higher path for CPI inflation pushes up the multiplier which is used to uprate the rateable values of non-domestic properties, raising receipts in the pre-measures forecast. The one-year extension of the relief for retail, hospitality, and leisure sectors (at 75 per cent generosity subject to a £110,000 cash cap) is expected to reduce receipts by £2.4 billion in 2024-25, while a freeze of the small business multiplier in 2024-25 is expected to further reduce receipts by around £0.4 billion a year from 2024-25.

Public sector expenditure

Summary of the expenditure forecast

4.40 Public spending as a share of GDP is forecast to fall marginally from 45.1 per cent of GDP in 2022-23 to 44.8 per cent of GDP in 2023-24, as the unwinding of energy support measures is largely offset by higher welfare costs. It is then forecast to continue to decline as a share of the economy to reach 42.7 per cent of GDP in 2028-29, largely due to falling departmental spending and debt interest. This means that, by the end of the forecast, public spending is 10.4 per cent of GDP lower than its post-war high of 53.1 per cent of GDP reached in the pandemic in 2020-21, but still 3.1 per cent of GDP higher than its pre-pandemic level of 39.6 per cent of GDP (Chart 4.7).

4.41 The changes in the spending-to-GDP ratio since our March forecast are:

• ONS revisions raised the level of nominal GDP by 1.9 per cent in 2022-23, which, before any other changes, would have reduced the spending-to-GDP ratio by 0.8 percentage points in each year of the forecast. The solid yellow line shows our March forecast restated for this higher GDP outturn.
• Our November 2023 pre-measures forecast has a lower spending-to-GDP ratio in 2022-23 due to lower outturn cash spending. Spending then rises above our restated March forecast by 0.4 percentage points on average over the forecast. This reflects higher forecast inflation and interest rates pushing up debt interest and welfare spending, more than offsetting lower departmental spending as a share of GDP.
• The direct and indirect effect of policies at this Autumn Budget raise spending marginally as a share of GDP (0.1 percentage points on average). In 2028-29, spending is 42.7 per cent of GDP.

Chart 4.7: Public expenditure as a share of GDP

4.42 The declining trajectory for public spending as a share of GDP over the next five years reflects downward paths for both department spending (‘DEL’), which is typically planned over multiple years and is set by the Government at Spending Reviews, and annually managed expenditure (‘AME’), which is less amenable to such multi-year planning (Table 4.4). In this forecast, the path of DEL beyond the current Spending Review period, which ends in 2024-25, is set by an assumption provided by the Treasury for the four years from 2025-26 to 2028-29. Both capital and current (‘resource’) DEL spending decline steadily as a share of GDP over the forecast, as nominal GDP grows more strongly than the path set by the Treasury’s spending assumption. Within AME, welfare spending rises, while debt interest and other AME fall over the forecast. The latter is driven by the end of energy support schemes, and lower spending on unfunded pensions and the EU financial settlement.

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

4.43 The ratio of public spending to GDP in 2028-29 is 3.1 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 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 3.3 per cent of GDP relative to our pre-pandemic expectations. Within this total, debt interest spending contributes 2.4 percentage points, due to upward revisions to both the amount of debt and the cost of servicing it. Revisions to welfare spending contribute 1.5 percentage points, due to higher inflation and increases in health- and disability-related benefit caseloads. Other factors reduce spending by 0.7 per cent of GDP.
• 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.
• Spending reductions announced in the 2022 Autumn Statement lower spending by 1.0 per cent of GDP in 2028-29, largely due to cuts to departmental spending.
• Spending increases announced in the March 2023 Budget raise spending by 0.3 per cent of GDP in 2028-29, largely reflecting more generous provision of free childcare hours and higher defence spending.
• In this Autumn Statement, spending in 2028-29 has been left broadly unchanged as an increase in the allocation to day-to-day departmental spending is offset by the assumption for flat cash departmental capital spending in 2028-29.

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

Changes in spending since our March 2023 forecast

4.44 Relative to our March forecast, spending in cash terms has been revised up significantly in each year of the forecast (Table 4.5). In the first half of the forecast, this increase is largely due to higher debt interest caused by higher RPI inflation and Bank Rate. Over the medium term, higher welfare spending accounts for the largest part of the change due largely to higher CPI inflation, while higher gilt rates mean debt interest spending remains significantly above our March forecast. Policies in this Autumn Statement raise spending by an average of £7.0 billion a year, driven by increases in resource DEL spending. And the indirect effects of the policy package raise spending by a further £1.7 billion a year on average, which is more than explained by the costs of servicing the additional debt issued to finance the overall fiscal loosening.

Table 4.5: Public sector expenditure: changes since March 2023

Table 4.6: Public sector expenditure as a share of GDP: changes since March 2023

Spending within departmental expenditure limits

4.45 Spending subject to departmental expenditure limits (DELs) makes up just over two-fifths 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.^[59] To facilitate comparison with our March forecast, throughout this section we exclude fiscally neutral additions to RDEL related to SCAPE rate changes (see Box 4.4 for an explanation of these changes). Our latest forecast (Tables 4.7 and 4.8) reflects:

• Departments’ detailed plans for 2023-24, as set out by the Treasury in July 2023 in its PESA publication,^[60] and departments’ plans for 2024-25, as announced in the October 2021 Spending Review, plus the effects of policy announcements since then. In these years, the Government has set department-by-department plans (Chart 4.9) and put in place tight spending controls.
• The Government’s post-Spending Review spending assumptions, which set overall spending totals, but not detailed plans, for 2025-26 onwards. These reflect assumptions about total departmental spending, as opposed to the department-by- department plans in the period covered by the Spending Review. 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.

Chart 4.9: Breakdown of total departmental spending in the forecast

Departmental spending in 2023-24 and 2024-25

4.46 Departmental resource spending in 2023-24 is expected to be £426.5 billion, up £15.0 billion (3.6 per cent) from 2022-23, £4.8 billion (1.1 per cent) higher than our forecast in March. In 2024-25, we expect departmental resource spending to rise to £430.1 billion, up £3.5 billion from March. The main changes compared to the March forecast are (Table 4.7):

• New funding for public sector pay increases, of £3.9 billion for the NHS in 2023-24, and £0.4 and £1.4 billion for other departments in 2023-24 and 2024-25, respectively.
• Other policy measures, including the spending impact of extending the retail, hospitality and leisure business rates relief cash cap to 2024-25 (£1.7 billion).
• Our judgement of reduced underspending (£0.9 billion in 2023-24 and 2024-25) to reflect the increase in inflation and wage pressures on departments’ cash-limited budgets since our March forecast.
• The reclassification of rail subsidies (-£1.7 billion in 2023-24 and -£1.0 billion in 2024-25), due to the reclassification of rail operating companies to public corporations. This is now included in our forecast as AME rather than DEL.

4.47 Departmental capital spending in 2023-24 is expected to be £96.0 billion, up £6.2 billion (6.9 per cent) on 2022-23 and broadly unchanged from our March forecast. In 2024-25, we expect CDEL spending to rise to £99.2 billion, up £2.3 billion (2.4 per cent) from March. The majority of this change is because we have lowered the assumed underspend in 2024-25 to £6.4 billion (from £8.3 billion in March) to reflect higher inflation pushing up the cost of capital projects and growth in capital budgets slowing after recent large increases. Policy changes since our March forecast add £1.8 billion on average, largely offset by other changes to CDEL limits.

Departmental spending from 2025-26 onwards

4.48 From 2025-26 onwards, we use the Treasury’s assumptions about the overall RDEL and CDEL envelopes.^[61]
At this Autumn Statement, RDEL grows by 0.9 per cent in real terms on average from 2025-26. This is lower in real terms than the 1.1 real terms growth in the RDEL envelope in every post-Spending Review year in our March forecast, but higher in cash terms due to a higher starting point in 2024-25 and higher forecast inflation. CDEL falls by £2.0 billion in 2025-26, then is held broadly flat in cash terms at the same level as March.

4.49 For the period beyond the Spending Review for which no detailed departmental spending plans exist, we have removed any assumed underspend against the RDEL and CDEL envelopes given to us by the Treasury. In our March forecast, we assumed RDEL and CDEL underspending would average £3.3 billion and £7.2 billion, respectively, from 2025-26 to 2027-28. We have made this change because:

• it is more transparent to treat the post-Spending Review assumptions as the expected actual level of spending, rather than assume it represents a limit against which departments will eventually underspend by some amount. The appropriate point to judge whether departments will underspend, and if so by how much, is when departmental allocations are set in Spending Reviews; and
• it is a more central view as departments have not systematically underspent against post-Spending Review envelopes in the recent past. In fact, recent history shows governments tend to increase envelopes as Spending Reviews approach – by over £30 billion at the last two Spending Reviews.^[62] So, looking at the balance of risks around our forecast, it is not central to assume that spending will eventually end up below any post-Spending Review assumption set by government.

4.50 The Treasury has chosen to exactly offset these changes within the forecast by reducing the DEL envelopes by an equivalent amount, making the overall change fiscally neutral.

4.51 Taking account of these revised assumptions and the policy changes announced in this Autumn Statement, RDEL averages £4.8 billion higher than our March forecast from 2025-26 to 2027-28. The main changes are:

• the update to the post-Spending Review assumption described above, which raises RDEL by £2.7 billion a year on average due to a higher starting level of RDEL in 2024-25 and higher forecast inflation, only partially offset by lower real growth on average than in March.
• an uplift through the post-Spending Review period reflecting public sector pay increases which averages £1.2 billion a year. The Treasury has chosen not to increase the post-Spending Review envelope for the NHS pay increases.
• other measures including labour supply (adding £0.7 billion a year on average) and business rates measures (reducing RDEL by £0.3 billion a year on average).

4.52 CDEL rises remains broadly flat in nominal terms at between £97 and £98 billion from 2024-25, which means it falls as a share of GDP from 3.5 per cent to 3.0 per cent over the same period. This reverses around half of the rise in CDEL as a share of GDP since 2019-20. Compared to our March forecast, CDEL is broadly unchanged in the post-Spending Review years. Our removal of assumed underspend in these years has no net effect on the forecast as the Treasury have lowered CDEL limits in response. In 2028-29, the Government has extended the cash freeze on CDEL, meaning it falls by £3.8 billion compared to our pre-measures forecast, which held it flat as a share of GDP.

Table 4.7: Departmental resource spending: changes since March 2023

Table 4.8: Departmental capital spending: changes since March 2023

4.53 Despite the increase in cash DELs described above, the real value of total DEL spending is lower in every year of the forecast as these cash increases are more than offset by higher forecast inflation. In March, we expected real DEL spending to increase by £12.8 billion between 2022-23 and 2027-28 (yellow line, Chart 4.10). On a pre-measures basis, our higher forecast GDP deflator inflation erodes the real value of DEL by £22.8 billion, resulting in real DEL spending falling by £10.0 billion over the same period (dashed blue line). Including the measures in this Autumn Statement, real DEL still falls – by £6.3 billion in 2027-28 (solid blue line). Maintaining the real value of DEL spending from our March forecast would therefore require DEL in 2027-28 to be higher by £19.1 billion. Box 4.3 examines the implications of this tighter real-terms departmental spending envelope for the Government’s post-Spending Review plans.

Chart 4.10: Change in real total DEL spending from 2022-23

Box 4.3: The Government’s post-Spending Review departmental spending plans

In the final four years of our forecast, departmental spending follows an overall envelope set by the Government because the detailed, department-by-department plans set in the 2021 Spending Review (SR21) only extend to 2024-25. For day-to-day spending (RDEL), this envelope grows by an average of 0.9 per cent a year in real terms. This is lower than the 1.1 per cent a year growth at our last forecast in March, though a higher starting point in 2024-25 and higher forecast inflation mean spending in cash terms is higher by £4.8 billion a year on average. For capital departmental investment spending (CDEL), the Government’s envelope implies total spending in cash terms is broadly unchanged since our last forecast. Taken together, these two assumptions imply total DEL falling as a share of GDP by 1.1 percentage points over the final four years of the forecast (though in 2028-29 DEL is still 1.2 percentage points above its pre-pandemic share of GDP). And by 2027-28, higher inflation means real total DEL spending in the same year is £19.1 billion lower than our March forecast.

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 D):

• 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.
• 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 then Chancellor Sunak’s statement that SR21 restored per pupil spending to 2010 levels in real terms.^c Prime Minister Sunak’s conference speech commitment to more teachers would need to be accommodated within this flat real spending per pupil and falling pupil numbers, to avoid further squeezing unprotected spending.
• Spending on the Government’s March 2023 Budget childcare policy to expand 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 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 4.1 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, for example the backlog in crown courts reached a record high of 65,000 in August 2023 and eleven ‘section 114s’ notices have been issued by local authorities since 2018, compared to two in the preceding 18 years. The Institute for Government’s recent report found that performance in eight out of nine major public services has declined since 2010, with schools the exception.^e 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 D: Implied post-SR21 breakdown of real RDEL spending and share of RDEL

Delivering these spending plans while maintaining or improving public services would require significant improvements in public sector productivity (i.e. producing the same outputs with fewer inputs). Measured public sector productivity fell sharply during the pandemic and, while it has recovered since, remains around 5 per cent below pre-pandemic levels. Raising public sector productivity by 5 per cent would be the equivalent of around £20 billion extra in funding, more than the £13.6 billion fall in ‘unprotected’ spending shown in Chart D. However, to turn such productivity improvements into lower spending, the Government would have to reduce spending on inputs, which would be likely to require dropping some of the input targets for staff and funding described above. An update to the Treasury’s Public Sector Productivity Programme, published alongside this Autumn Statement, has identified areas for potential future productivity improvements, such as by using AI and cutting administrative tasks for frontline workers.

Overall, the Government’s post-Spending Review plans present a significant risk to our forecast. As recent Spending Reviews have approached, governments have topped up annual day-to-day spending envelopes significantly: by £39 billion (14 per cent) on average in the year up to the November 2015 Spending Review, and by £32 billion (8 per cent) in the October 2021 Spending Review. 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

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 2022-23 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

4.54 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.

4.55 Welfare spending is forecast to rise sharply this year (by £33.9 billion, or 13.0 per cent) and next (by £21.1 billion, or 7.1 per cent), driven by the uprating of most benefits with CPI inflation. Benefits were uprated by 10.1 per cent in April this year and are expected to be uprated by 6.7 per cent, the September 2023 CPI inflation figure, in April 2024. Welfare spending is then forecast to rise by an average of £11.3 billion (3.4 per cent) a year over the rest of the forecast period. The main drivers are health and disability benefits (reflecting rising caseloads for these benefits), and pensioner spending (due to the ageing population and the triple lock). Spending on these components explains nearly nine-tenths of the increase in total welfare spending between 2024-25 and 2028-29.

Table 4.9: Total welfare spending

4.56 Spending on universal credit (UC) and its legacy equivalents is now forecast to rise notably, by 39.8 per cent (£29.2 billion) from 2022-23 to 2028-29, of which around two-thirds is due to uprating. Decomposing UC and legacy spending into its five key components^[63]
the largest growth is in (Chart 4.11):

• Health-related spending^[64] which increases by £12.0 billion (55 per cent) over the forecast due to rising caseloads from worsening health trends and rising take-up, accounting for two-fifths of the total rise in UC and legacy spending across the forecast. The number of claimants in receipt of these health-related benefits rises by 0.6 million, from 2.8 million in 2022-23 to 3.4 million in 2028-29.
• Housing spending increases by £9.6 billion (41 per cent) over the forecast, explained by an increasing take-up rate of the housing element among UC claimants.
• Unemployment-related spending^[65]increases by £4.1 billion (108 per cent) over the forecast, due to increasing levels of unemployment and the increase in the administrative earnings threshold (AET) to 18 hours in April 2024. The AET change increases the number of hours a claimant can work while remaining in DWP’s ‘intensive work search’ regime, which moves a further 135,000 UC claimants into the unemployment-related caseload when it is introduced in 2024-25.
• Child spending^[66]increases by £2.7 billion (17 per cent) over the forecast as the number of claimants with disabled children, who qualify for the additional disabled child payment, increases.
• Other UC and legacy spending^[67] increases by £0.8 billion (9 per cent) over the forecast, with upward pressure from the rising UC in-work caseload and higher take-up of the carer element offset by increasing savings due to the earnings taper and other deductions.

Chart 4.11: Universal Credit and legacy spending by component

4.57 Relative to our March 2023 forecast, welfare spending is now expected to be higher in all years by an average of £13.5 billion. Table 4.10 shows that this revision is driven by:

• Higher uprating due to increases in our inflation and earnings forecasts. This increases spending by an average of £10.0 billion over the forecast, and accounts for around three-quarters of the total upwards revision since March 2023.
• Higher unemployment, which increases jobseeker’s allowance and UC caseloads, increases spending by an average of £1.6 billion over the forecast.
• A downward revision to our forecast for incapacity and disability caseloads, which reduces spending by an average of £0.9 billion over the forecast period. While outturn data continues to show that new claims for PIP and UC health journeys have been increasing, we now assume that this growth is in part driven by the current cost-of-living pressures and therefore that caseload growth rates will slow as real household disposable incomes recover in the second half of our forecast.
• Other forecast changes, which add £2.2 billion on average over the forecast, including higher-than-expected caseload outturn data for UC (£0.9 billion), revisions to the profile of fraud and error (£0.6 billion), and a higher housing benefit forecast in light of higher-than-expected outturn (which adds £0.5 billion a year).
• Policy measures in this Autumn Statement, which increase spending by an average of £0.6 billion over the forecast. The largest cost is increasing the local housing allowance, which costs an average of £1.2 billion a year over the forecast – and is only partially offset by the policy to reform the work capability assessment gateway, which saves an average of £0.6 billion a year over the forecast.

Table 4.10: Welfare spending: changes since March 2023

Locally financed expenditure and public corporations’ expenditure

4.58 We expect locally financed current expenditure to rise from £62.2 billion in 2023-24 to £75.3 billion in 2028-29, as local sources of income grow steadily by an average of £2.6 billion a year.^[68] Compared with the March 2023 forecast, locally financed current expenditure is £1.5 billion a year higher on average (Table 4.11):

• Net use of current reserves increased by £2.3 billion in 2022-23. This is the first time since 2019-20 that local authorities have drawn on their reserves for current spending. Due to the ongoing funding pressures on local authorities, we assume that there will be further drawdowns during the current Spending Review period, of £1.5 billion in 2023-24 and £0.8 billion in 2024-25, compared to an assumption of no drawdown in both years in our March forecast.
• The other source higher current spending is stronger council tax of £0.5 billion on average, due to higher outturn in 2023-24 than forecast in March 2023.
• Policy at this Autumn Statement to extend the retail, hospitality and leisure business rates relief cash cap to 2024-25 reduces retained business rates by £1.5 billion. From 2025-26 to 2028-29, another policy for some combined authorities to retain 100 per cent of business rates increases current expenditure by £0.7 billion per year.

4.59 Since 2018, there have been eleven ‘section 114s’ notices issued by local authorities, compared to the two issued since 2000.^[69] These notices indicate that the authority’s forecast income is insufficient to meet its forecast expenditure for the next year.^[70] Some of these have been issued due to unique financial management issues and some local authorities have submitted multiple times. The direct impact on our forecast to date has been relatively small as the central government response to section 114s has been to allow affected local authorities to reallocate their capital budget towards day-to-day spending (a ‘capitalisation direction’) or to increase council tax rates.

4.60 However, there is still a risk to our forecast from wider pressures on local authority finances. Since 2010-11, local authority spending has fallen from 7.4 to 5.0 per cent of GDP, and it falls further in our forecast to 4.6 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.

4.61 Locally financed capital expenditure is expected to fall slightly in 2023-24 to £8.0 billion and then to £7.7 billion in 2028-29.^[71] Borrowing for capital expenditure is expected to fall from its 2019-20 peak (£11.5 billion) to £7.0 billion in 2028-29. This reflects the financial pressures facing local authorities and higher interest rates on loans from the Public Works Loan Board, their principal source of financing. Public corporations’ capital expenditure is expected to grow steadily from £11.0 billion in 2023-24 to £12.1 billion in 2028-29. Compared to the March 2023 forecast, we have incorporated latest ONS outturn that has reduced public corporations’ expenditure by £0.3 billion a year on average.

Table 4.11: Locally financed and public corporations’ expenditure: changes since March 2023

Debt interest spending

4.62 Debt interest spending, net of Asset Purchase Facility (APF) flows, is forecast to be 4.3 per cent of GDP in 2023-24, 0.6 percentage points higher than forecast in March. This would be the second highest level since the Second World War after 2022-23 (Chart 4.12). The increase from March reflects the 3.8 percentage point upward revision to our forecast for RPI inflation in 2023-24, in combination with a 0.9 percentage point upward revision to our forecast for Bank Rate.^[72]

4.63 Debt interest spending is then forecast to fall back somewhat to 3.8 per cent of GDP and 9.2 per cent of total revenue in 2028-29. Compared to our March forecast, it is on average 1.6 percentage points higher as a share of total revenue and 0.6 percentage points higher as a share of GDP between 2023-24 and 2027-28.

Chart 4.12: Debt interest spending relative to GDP and revenues

4.64 In nominal terms, debt interest spending reaches £116.2 billion in 2023-24, before falling to £101.9 billion in 2025-26 and then rising year-on-year to peak at £122.5 billion in 2028-29 (Table 4.12). It has been revised up in each year of the forecast relative to March, reflecting several factors:

• Our forecast for RPI inflation is higher by 3.8 percentage points this year but by declining amounts thereafter. This pushes up the cost of index-linked gilts and accounts for £15.5 billion (70 per cent) of the upward revision in 2023-24, but only £1.7 billion (9 per cent) by 2027-28. We expect the impact of higher RPI inflation in 2023-24 to push up borrowing compared to our March forecast in the second half of the fiscal year.
• Higher Bank Rate in every year of the forecast, which increases spending by an average of £7.5 billion a year (with around four-fifths of this relating to the APF). The upward revisions to Bank Rate decline gradually from 2025-26 onwards but remain 1.1 percentage points higher in 2027-28.
• Higher gilt rates, which account for an increasing share of the upwards revisions and drive £11.2 billion (three-fifths) of the increase by 2027-28, as the weighted-average conventional gilt rate is 110 basis points above our March forecast.
• Other changes, including to the Government’s net financing requirement, which reduces spending in the first three years, but raises it slightly thereafter.

Table 4.12: Central government debt interest (net of APF): changes since March

Unfunded public service pensions

4.65 The net cost of unfunded public service pensions (outlays minus contributions) is forecast to fall from £5.6 billion in 2023-24 to £0.8 billion in 2028-29. Relative to our March forecast, spending has been revised down by an average of £5.0 billion a year due to:

• The new SCAPE discount rate, to be implemented in April 2024, increases the present value of future pensions payments and in turn causes employer contributions to rise. This reduces net pensions spending by an average of £23.6 billion a year. Box 4.4 explains the SCAPE rate change and its overall impact on the public finances.
• Other changes resulting from pension schemes’ ongoing valuations reduce employer contribution rates and so increase spending by £18.0 billion a year on average, largely offsetting the impact of the SCAPE rate. These changes are driven by slower improvements in life expectancy (based on the latest ONS estimates) and higher-than-expected contributions since the last valuation meaning lower contributions are required in the future.
• Other forecast changes are relatively small in comparison, lowering spending by £0.4 billion a year on average compared to March. This reflects stronger pensionable earnings growth pushing up receipts across the forecast, reducing spending, which more than offsets higher CPI inflation increasing expenditure.

Box 4.4: The fiscal implications of SCAPE rate changes

Unfunded public service pensions include central government pay-as-you-go schemes (the largest of which are the Civil Service, National Health Service, Teachers, and the Armed Forces) and locally administered police and firefighters’ schemes. Public spending on these schemes is measured in AME in net terms as total payments to each scheme’s pensioners less total pension contributions from public sector employers and current employees.

The Superannuation Contributions Adjusted for Past Experience (SCAPE) rate is the discount rate used by unfunded public service pension schemes to place a current value on expected future pension payments. It has been subject to a series of adjustments over the past eight years, reducing the SCAPE discount rate from CPI plus 3 per cent in 2011 to CPI plus 2.4 per cent in 2018. In the most recent update, the Government announced a further decrease to CPI plus 1.7 per cent, which will impact employer contribution rates from April 2024, prompted by revisions to our long-term growth forecasts.

A lower discount rate means that the present value of past and future pension promises to public sector workers is higher than previously estimated. This results in a higher valuation of existing pension liabilities and the cost of ongoing pensions being accrued by current members of the schemes. This increase in liabilities and the higher expenditure needed to cover these costs might be expected to increase measures of both debt and the deficit. However, the accounting rules used by the ONS for the main fiscal aggregates combined with the operation of departmental spending limits means this is not the case, which is an example of a ‘fiscal illusion’.^a Table A shows the impact of SCAPE-related changes in this forecast:

• Net pensions AME spending: The ONS account for pension expenditure in cash terms within AME. Cash payments to current pensioners made by pension schemes are not affected by the SCAPE change. However, pension scheme receipts increase, as the SCAPE change leads to an increase in employer contributions required to finance the higher future liabilities. Overall, therefore net unfunded pension spending in AME falls. In this forecast, this has been partially offset by the other changes described in paragraph 4.65.
• DEL spending: The net effect of the changes in AME require public sector employers to increase contributions by an average of £5.6 billion per year. The Government has committed to providing funding for these increased costs which raises DEL spending to exactly offset the impact.^b
• PSNB: The net result of these changes in AME and DEL spending means there is no impact on PSNB.^c Payments out of the public sector to pensioners are unchanged, while higher employer contributions have been offset by the addition to DEL.
• PSND: Pensions liabilities are not recognised in PSND. So, the stock change in liabilities caused by the SCAPE change has no direct impact on PSND.
• PSNW: Pensions liabilities are recognised in PSNW but changes in the SCAPE discount rate will not impact the liability recorded by the ONS as they use a constant discount rate.

Table A: Impact of SCAPE rate changes in our forecast

The reduction in the SCAPE rate reflects our more pessimistic view on long term economic growth as set out in the 2022 Fiscal risks and sustainability report. All else equal, this reduces the expected future tax revenues available to pay for future pension payments. So, in isolation, the SCAPE rate change suggests the public finances are less sustainable than previously thought. But the overall sustainability of unfunded public pensions is also affected by numerous other factors, including the rate of future pay increases which are also likely to be affected by lower growth. We will address these wider issues when we update our comprehensive estimates of the future fiscal impacts of pensions as part of our next long-term fiscal projections in summer 2024.

a) For more detail on the treatment on unfunded pension schemes in the National Accounts, see National Audit Office, The cost of public service pensions, 2010.

b) Once all schemes valuations are complete and final DEL funding is confirmed by the Treasury, there are likely to be some relatively small PSNB impacts as a result of devolved schemes and Barnett consequentials, and contributions in respect of employers not fully covered by the Government’s funding commitment for centrally funded employers.

c) If the Government had not increased DEL spending to compensate departments for higher employer contributions, there would be a reduction in PSNB, as departments would have been required to fund increased contributions from within their existing budgets so reducing funding available for other spending.

Other annually managed expenditure

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

• Scottish Government current expenditure has increased by an average of £1.9 billion a year over the forecast, due to higher RDEL spending and tax receipts.
• Spending on domestic energy schemes in 2022-23 was £2.8 billion lower in outturn than our forecast in March. For non-domestic energy schemes, we have revised down spending in 2023-24 by £0.5 billion to £0.6 billion, based on lower-than-expected outturn data in the first half of the year.
• Funded pensions’ current spending has been revised down £3.7 billion on average a year. This is driven by taking on the latest ONS outturn data, which implies an improved funding position of these pension schemes over recent years.
• Company tax credits have been revised down by £0.5 billion on average per year. This is driven by a combination of lower outturn spending and slower growth in our forecast for investment.

Deficit aggregates

4.67 Public sector net borrowing is forecast to be £123.9 billion (4.5 per cent of GDP) this year having fallen from a post-war high of £311.9 billion (15.0 per cent of GDP) during the pandemic in 2020-21 Chart 4.13. It is then expected to fall over the forecast period to £35.0 billion (1.1 per cent of GDP) in 2028-29, the final year of the forecast, which would be the lowest as a share of GDP since 2001-02.

Chart 4.13: Public sector net borrowing

4.68 As shown in Chart 4.14, this trajectory is driven by a steady rise in receipts as a share of GDP combined with a fall in overall spending. A large contribution to the increase in receipts is rising income tax and NICs due to high inflation and nominal earnings growth combined with frozen thresholds. A further large contribution comes from departmental and other spending falling as a share of GDP. Debt interest also contributes to the decline due to a fall in interest costs as inflation abates across the forecast period, while welfare spending pushes up on borrowing due to high inflation and rising caseloads.

Chart 4.14: Public sector net borrowing: contribution to fall from 2023-24 to 2028-29

4.69 In cash terms, the changes in borrowing since our March 2023 forecast are shown in Chart 4.15 and Table 4.13. Changes to our underlying forecast, (white diamonds) before the impact of new policy, reduce borrowing by £15.8 billion in 2023-24 and £14.5 billion in 2024-25 rising to £26.8 billion in 2027-28. These changes reflect:

• Improvements in our pre-measures receipts forecast, (yellow bars) which reduces borrowing by £35.3 billion in 2023-24 rising to £59.0 billion in 2027-28. This is largely a result of higher forecast inflation and earnings which drive higher income tax, NICs, corporation tax and VAT. Overall, higher inflation increases receipts and reduces borrowing by £53.5 billion by 2027-28, and a higher effective tax rate, mainly reflecting more fiscal drag, reduces borrowing by a further £27.1 billion in 2027-28. This is partly offset by a lower forecast for real GDP reducing receipts and increasing borrowing by an average of £21.6 billion.
• The increase in receipts is partially offset by higher pre-measures spending (blue bars) in each year of the forecast, by £19.5 billion in 2023-24, £33.3 billion in 2024-25 and an average of £33.0 billion thereafter. Higher debt interest spending accounts for most of the upward revisions between 2023-24 and 2025-26 due to higher to RPI inflation, Bank Rate and gilt rates. Welfare payments are also higher across the forecast, reflecting increased caseloads and the cost of uprating benefits, due to higher CPI inflation.

4.70 Changes in borrowing relating to policy decisions
announced in this Autumn Statement (green bars and shown in Table 4.13) increase borrowing by an average of £17.0 billion a year, relative to March. This reflects:

• The direct effect of policy decisions, which increase borrowing every year. In 2023-24 new measures cost £8.7 billion, rising in each subsequent year to reach £25.1 billion in 2027-28 before falling to £17.3 billion in 2028-29. The cuts to NICs increase borrowing by an average of £8.6 billion from March. The permanent full expensing measure increases borrowing by an average of £9.1 billion in the final two years. Other receipts policy decisions reduce borrowing from March by an average of £1.3 billion. Spending decisions cost an average of £7.1 billion per year, which includes an increase in departmental spending (DEL) of £5.3 billion per year on average.
• The indirect effect of policy decisions, which reduce borrowing in the near term but increase it by the end of the forecast period. The policy package temporarily boosts demand in the near term and leads to a relatively small, but lasting, increase in supply. This reduces borrowing with the impact peaking at £2.7 billion in 2025-26. However, by the end of the forecast this is more than offset by the debt interest costs of the additional borrowing for the policy measures.

4.71 Overall, our forecast for post-measures borrowing in 2027-28 is broadly unchanged from March 2023 as the Chancellor has spent virtually all of the cumulative £88.5 billion pre-measures improvement in the forecast on his Autumn Statement measures. As explained above, this pre-measures forecast improvement is primarily driven by higher inflation and earnings boosting tax receipts. While some of this is offset by inflation also driving higher welfare and debt interest spending, the Chancellor has only increased DEL spending by £5.3 billion on average compared to March. As set out in paragraph 4.53 if DEL spending had instead been increased to maintain its real value, it would have increased borrowing by £19.1 billion in 2027-28.

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

Table 4.13: Public sector net borrowing: changes since March 2023

4.72 As is outlined in the note of Table 4.13 there are elements of receipts and spending that have no impact on public sector net borrowing as they have offsetting effects. Compared to March, changes to these elements have led to a sizeable increase in both receipts and spending by an average of £8.9 billion across the forecast. This includes the pre- and post-measures effects of SCAPE pension changes (see Box 4.4) where the impacts on underlying spending are wholly offset by policy decisions. There is also an increase in spending on environmental subsidies due primarily to lower forecast electricity prices leading to increases in the expected cost of subsidies for renewable electricity generation. These are fiscally neutral because they are offset by increased levies on electricity bills.

Table 4.14: Public sector net borrowing neutral receipts and spending: changes since March 2023

Other deficit aggregates

4.73 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’. 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 the position in the economic cycle, which provides an indication of the underlying or structural deficit.

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

• The primary deficit is 1.0 per cent of GDP this year, however it moves into a surplus of 0.2 per cent of GDP in 2025-26, which rises to 2.2 per cent of GDP by 2028-29. The primary balance has not been in surplus since 2001-02.
• The current deficit falls from 1.9 per cent of GDP this year to surplus a of 0.3 per cent of GDP in 2027-28, rising to 0.7 per cent of GDP in 2028-29.
• Cyclically adjusted measures of the deficit are higherthan unadjusted metrics in 2023-24, but lower in each subsequent year. This reflects the output gap which initially widens, as spare capacity in the economy increases in each year until 2025-26, but diminishes thereafter and closes by 2028-29.

Financial transactions including the Asset Purchase Facility

4.75 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. Valuation effects capture changes in the value of the assets held by the public sector.

4.76 Financial transactions and valuation effects increase headline debt by £39.8 billion in 2023-24 and by progressively smaller amounts in each year until 2025-26, when they reduce it by £50.3 billion, before increasing debt in each year until 2028-29, when they add £42.7 billion. The uneven effects on debt are driven largely by repayments of Bank of England Term Funding Scheme (TFS) loans. Absent the TFS, financial transactions add to debt in all years by an average of £46.0 billion, reflecting the losses forecast on the gilts sold by the Asset Purchase Facility (APF) and transactions relating to student loans.

4.77 Revisions since our March forecast add to headline debt in each year, by amounts which peak at £25.8 billion in 2025-26 but fall to £14.1 billion by 2027-28. This reflects several factors:

• Additional early repayments of TFS loans which, relative to our March forecast, reduce debt by £28.1 billion in 2023-24 (reflecting the £17.0 billion of repayments made so far this year), but mean fewer repayments later in the forecast.
• Payments for funded pension schemes, which we have revised by an average of £4.7 billion and now add to debt in each year. This reflects the higher-than-expected outturn for 2022-23, which raises cash spending in each year of the forecast.
• Revisions to our forecast for student loan interest receipts which add a maximum of £3.7 billion to debt in 2025-26 and an average of £1.8 billion a year, reflecting the upward revisions to our forecast for RPI inflation.

4.78 Relative to our March forecast, contributions from APF-related flows to headline debt have risen by an average of £11.2 billion. Across the forecast the cumulative impact of the APF on headline debt is £125.1 billion, of which over three-fifths is via interest losses that affect PSNB. Interest losses have increased almost eight-fold in 2023-24 relative to 2022-23 as Bank Rate increased to a 15-year high. Box 4.5 discusses the impact of the APF on key fiscal aggregates in further detail.

Box 4.5: Fiscal accounting for quantitative easing and tightening

The Asset Purchase Facility (APF) houses the assets purchased by the Bank of England as part of its programme of quantitative easing initiated in 2009. The Treasury receives the profits from the APF and also indemnifies the Bank against any losses from it. Excess cash held in the APF started to be transferred to the Treasury in January 2013. Between then and July 2022, when interest rates were at historically low levels, the Bank of England transferred £123.9 billion of cash profits on the APF to the Treasury. Since October 2022, as interest rates have risen and so the APF has started to incur losses, £29.1 billion has been transferred from the Treasury to the APF. We expect continued losses into the future. These result from both interest losses and the crystallisation of valuation changes. These hit fiscal aggregates in different ways:

• Public sector net borrowing records the interest losses that occur when the interest received on holdings of gilts and other assets is less than the interest paid on central bank reserves.
• Public sector net debt also records the interest losses. In addition, as stocks of APF assets are sold, any differences between the redemption price at which they are booked in the public finances and the sale price results in a change in the recorded level of debt.
• Public sector net debt excluding the Bank of England is impacted as the APF calls on the Treasury to make good on losses. These include both losses from interest and the difference between original purchase price and the eventual sale or redemption price.

Chart A shows how we expect the costs associated with the APF to evolve over its lifetime. Our updated estimate reflects the rise in Bank Rate and gilt rates since March as well as a change in our APF runoff assumption. Our March forecast assumed a constant £80 billion decline in its size each year. We now assume a £100 billion decline between October 2023 and September 2024 reflecting the Monetary Policy Committee’s stated intention.^a This is followed by an assumption of a constant pace of active sales of £48 billion a year, so that the overall pace of APF reduction varies year-to-year in line with the uneven profile for redemptions across the forecast period.^b This follows guidance issued by the MPC in August 2023 that ‘sales must be conducted in a relatively gradual and predictable manner over a period of time. The focus of the MPC is on total gilt stock reduction, comprising both maturing gilts and sales, such that natural variation in maturities may also lead to variation in the pace of sales over time. Notwithstanding that, the Committee places some weight on continuity in sales’.^c

Current market expectations for Bank Rate and gilt rate, combined with the latest runoff path, imply a cumulative net lifetime loss of £126.0 billion, which is £63.4 billion higher than we estimated in our July Fiscal Risks and Sustainability 2023 report (based on our assumptions from March) due to the higher path for Bank Rate and gilt rates. Our new runoff assumption concentrates these losses over a shorter period, meaning the APF also reaches its lifetime cost one year earlier than we assumed in July.

Chart E: Forecast of cumulative flows to and from the APF

These estimates of the lifetime cash flow of the APF are highly uncertain and highly sensitive to the assumptions made in the calculations, differing from previous Bank of England estimates of this value. It is also important to stress that this narrow summary of the lifetime cashflows associated with QE and quantitative tightening (QT) is not an assessment of the overall fiscal (let alone economic) impact of the QE 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 taken into account in any comprehensive assessment of the impact of QE.

a) See Bank of England Monetary Policy Summary and minutes of the Monetary Policy Committee meeting, September 2023

b) Our £48 billion active sales assumption is based on the average of the active sales undertaken in the year to September 2023 and plans for the year to September 2024.

c) See Bank of England Monetary Policy Report, August 2023.

4.79 The central government net cash requirement (CGNCR, the amount of cash the government needs to raise each year), falls from £150.7 billion in 2023-24 to £93.9 billion by 2028-29, generally following the path for the deficit. This will largely be financed by gilt issuance which, alongside the unwinding of APF gilt holdings by the Bank of England, means the private sector needs to absorb historically high volumes of debt over the coming years.

4.80 Chart 4.16 shows that in the five years preceding the financial crisis and the advent of QE, gross gilt issuance (the nominal amount of debt issued by the government) was around 3.9 per cent of GDP and net issuance (gross issuance net of redeeming debt) averaged 2.3 per cent of GDP. Over the forecast we estimate these averages will rise to 8.4 per cent of GDP and 3.5 per cent GDP respectively. As a result, we estimate that the change in private sector holdings will reach its highest level on record of 7.9 per cent of GDP in 2024-25 and an average of 6.4 per cent of GDP over the forecast period, the highest sustained level this century and 0.8 per cent of GDP a year higher than our March forecast.^[73]

Chart 4.16: UK gilt issuance and change in private holdings since 1999-00

Debt and other balance sheet aggregates

4.81 Public sector net debt excluding the Bank of England (PSND ex BoE) rises from 89.0 per cent of GDP in 2023-24 to 93.2 per cent in 2026-27. It then declines in the final two years to 92.8 per cent of GDP by 2028-29. Relative to our March forecast it is lower in every year, by an average of 2.1 per cent of GDP, due mainly to upward revisions to nominal GDP which are only partly offset by higher cash debt.^[74]

4.82 On a pre-measures basis, cash debt is £21.8 billion higher in 2027-28 compared to March, reflecting higher financial transactions more than offsetting the forecast reduction in pre-measures borrowing. The cost of the measures announced at this Autumn Statement means that post-measures cash debt is in total £106.7 billion higher by 2027-28 than in the March forecast.

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

Table 4.15: Public sector net debt (excluding Bank of England): changes since March

4.83 Public sector net debt (PSND), the headline measure of debt, rises from 97.9 per cent of GDP in 2023-24 to a peak of 98.6 per cent of GDP in 2024-25, before falling in every year to 94.1 per cent of GDP in 2028-29. Alongside PSND and PSND ex BoE, we also produce forecasts for public sector net financial liabilities (PSNFL) and public sector net worth (PSNW) – broader balance sheet measures shown in Chart 4.18. These provide a more comprehensive picture of the public sector balance sheet than PSND.

4.84 While underlying debt rises by 3.8 per cent of GDP between 2023-24 and 2028-29, PSNFL and (inverted) PSNW both fall, by 4.9 per cent of GDP and 9.9 per cent of GDP respectively. The fall in PSNFL reflects assets which are not recorded within debt, of which around two-thirds relates to student loans. (Inverted) PSNW improves faster still, reflecting a 2.9 per cent of GDP increase in non-financial assets and a 1.9 per cent of GDP fall in unfunded pension liabilities.

Chart 4.18: Four measures of the public sector balance sheet

Risks and uncertainties

4.85 Over recent years, large shocks and their aftermath have resulted in significant revisions to our economic and fiscal forecasts from one fiscal event to the next. The same is true in this forecast with the net effect of historically large changes in inflation, wage growth, and interest rates since March proving to be a net positive for the public finances. This led to a significant reduction in borrowing on a pre-measures basis which the Chancellor has chosen to spend virtually all of on his policy measures. However, history shows there is a risk of significant changes in the forecast at future events, which may be less favourable.^[75]

4.86 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. To illustrate three key uncertainties in this forecast:

• The tax-to-GDP ratio is forecast to increase to a post-war high by the forecast horizon. This is driven by judgements in our economy forecast such as continued relatively high nominal growth in earnings and consumption. It is also partly driven by our judgement that much of the higher-than-expected receipts outturn seen so far in 2023-24 reflects higher effective tax rates that will in part be maintained throughout the forecast. These are our central judgements based on the latest public finances data and economic conditions, but they could prove to be too optimistic or too pessimistic. If the tax-to-GDP ratio remained at its 2023-24 level, tax revenues would be £44.7 billion lower in 2028-29.
• With a debt stock of almost 100 per cent of GDP, the public finances remain very sensitive to interest rates, as shown in this forecast where both Bank Rate and gilt rates have been revised upwards significantly since March, which has increased debt interest spending by an average of £14.1 billion. We estimate that a 1 percentage point increase in both Bank Rate and gilt rate in every forecast year would increase debt interest spending by £18.6 billion in 2028-29.
• The number of working-age people classed as economically inactive has risen by over 400,000 since the beginning of the pandemic which is more-than accounted for by those citing long-term sickness as the reason. This is an important trend driving our forecast for the participation rate which is a key driver of both welfare spending and taxes on earnings. However, it is very uncertain and is further complicated by the suspension of ONS Labour Force Survey (LFS) estimates in September and October 2023 due to quality concerns from falling survey response rates.^[76] In our 2023 Fiscal risks and sustainability report, our downside scenario for health-related inactivity estimated that borrowing could be £21.3 billion higher in 2027-28.

4.87 In Chapter 5, we also illustrate risks to the forecast through scenarios considering two other key economic drivers of the forecast:

• Inflation can have a significant impact on the public sector finances as has been illustrated in this forecast. In our scenarios we explore the different impacts on the public finances of imported inflation versus domestically generated inflation. Imported inflation generally leads to higher borrowing due to higher debt interest and welfare spending. Domestically generated inflation can be positive for the public finances as it also boosts tax receipts, though this improvement largely relies on an erosion of the value of real departmental spending. If the government responded to higher inflation by maintaining the real value of spending, most of the improvement would disappear.
• Productivity growth is currently forecast to be above the post-financial crisis rate but still well below the rate seen before the financial crisis of around 2 per cent. Productivity is the key driver of economic growth – and therefore the outlook for the public finances – over the medium and long-term. In the scenarios, we estimate the impact of productivity being 0.5 percentage points above or below our central estimate and show that it would reduce borrowing by 46.0 billion or increase it by £42.2 billion, respectively, by the end of our forecast.

4.88 While this Autumn Statement crystalises one policy risk by making temporary full expensing permanent at a cost of £9.2 billion in 2028-29, there remain a range of policy-related risks to the outlook.

• For example, the Government’s stated policy is to increase the fuel duty rate in March 2024 after the latest one-year freeze announced in the Spring Budget. This would be the first time any Government has raised fuel duty rates in cash terms since 2011 and would increase the price of petrol and diesel by around 8 pence a litre. Not going ahead with this and freezing the duty at its current rate would reduce government revenue by £6.2 billion in 2028-29. We have calculated the contribution of RPI indexation of fuel duty rates to the Chancellor’s headroom against his fiscal targets in Chapter 5 (see Table 5.1).
• Departmental expenditure limits (DELs) after the current Spending Review period also present a significant policy risk to the forecast. As previous Spending Reviews have approached, governments have topped up annual day-to-day spending envelopes significantly by £39 billion (14 per cent) on average in the year up to the November 2015 Spending Review, and by £32 billion (8 per cent) in the October 2021 Spending Review. 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.

4.89 As set out in Box 4.3 the Government has stated ambitions to increase spending in key areas which are not captured in our forecast. The Government has a stated aspiration to spend 2.5 per cent of GDP on defence at an unspecified point in the future, as fiscal and economic circumstances allow. Relative to the NATO minimum of 2 per cent of GDP, the cost of meeting this commitment if it were met in 2028-29 would be around £16 billion. The Treasury also expect a return to the 0.7 per cent of GNI spend on Official Development Assistance (ODA) post-2027/28 after it fell to 0.5 per cent of GNI from 2021 as a temporary measure due to the impact of the pandemic.^[77] Returning to the previous target would cost around £6.3 billion if it were met in 2028-29.

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 5.17).

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 modest margins in the new fifth year of our forecast (2028-29), and the welfare cap is on course to be missed in 2024-25:

• The fiscal mandate for public sector net debt (excluding the Bank of England) is met by a margin of 0.4 per cent of GDP (£13.0bn) in 2028-29. This is a 0.2 per cent of GDP (£6.5 billion) improvement relative to the headroom of 0.2 per cent of GDP (£6.5 billion) against getting debt falling in 2027-28, the fifth year of our March 2023 forecast.
• The 3 per cent of GDP borrowing target is met by a margin of 1.9 per cent of GDP (£61.5 billion in 2028-29), a 0.6 per cent of GDP (£22.3 billion) improvement relative to our March forecast for headroom of 1.3 per cent of GDP (£39.2 billion) in 2027-28.
• The welfare cap in 2024-25 is on course to be exceeded by £8.6 billion, a £4.5 billion deterioration relative to the £4.1 billion by which it was missed in March.

5.5 Given the Treasury Select Committee’s concerns about our forecasts assuming rarely implemented RPI indexation of fuel duty rates,^[78] 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 £6.2 billion lower than our central forecast in 2028-29, so the fiscal mandate to get debt falling would still be met, but debt would only fall by 0.2 per cent of GDP (£7.4 billion), removing 43 per cent of the Chancellor’s headroom, while the borrowing target would be met by £55.4 billion (1.7 per cent of GDP).^[79]

Table 5.1: Performance against the Government’s fiscal targets

Change in headroom against fiscal targets

Debt falling and public sector net borrowing targets

5.6 Improvements in our pre-measures fiscal forecast increase headroom against both the debt falling and borrowing targets relative to our March 2023 forecast. The rolling nature of both targets gives the Government an extra year to meet both targets, which further increases pre-measures headroom in the new target year of 2028-29 relative to the previous target year of 2027-28. Policy measures announced in the Autumn Statement reduce the headroom. As shown in Chart 5.1:

• Underlying forecast changes add £19.6 billion and £31.0 billion respectively to the headroom against the debt falling and borrowing targets in 2027-28. For the debt target, the increase in headroom compared with March is largely explained by the lower borrowing in cash terms in 2027-28, which is partly offset by higher financial transactions (reflecting higher losses on the sale of gilts from the Asset Purchase Facility (APF)).
• The rolling nature of the fiscal mandate and the supplementary target means that the relevant year for these targets is 2028-29 compared with 2027-28 in March. Rolling onto the new target year boosts headroom by £4.9 billion for the fiscal mandate and by £10.4 billion for the supplementary target on a pre-measures basis. This reflects the fact that borrowing is lower in 2028-29 compared with 2027-28, due to a further year of nominal receipts growing faster than GDP, while spending only grows slightly in cash terms and continues to fall as a share of GDP in 2028-29. And, for the fiscal mandate, the increase in headroom is also a result of the higher starting level of cash debt.
• Autumn Statement policy measures reduce the headroom against both targets by £18.0 billion (for the fiscal mandate) and £19.1 billion (for the supplementary target), offsetting between a half and three-quarters of the pre-measures improvement in headroom.
• This leaves overall headroom at £13.0 billion (0.4 per cent of GDP) against the debt falling target and £61.5 billion (1.9 per cent of GDP) against the target for borrowing to not exceed 3 per cent of GDP.

Chart 5.1: Fiscal target headrooms: changes since March

5.7 Much of the change in headroom relative to our March forecast reflects the impact of higher inflation on the public finances. As outlined in Box 4.1, of the £57.0 billion increase in pre-measures tax receipts in 2027-28 around £50.9 billion can be attributed to higher inflation.^[80]
This is only partly offset by the effects of inflation on annually managed expenditure (AME), which raises welfare and debt interest spending by £27.4 billion.^[81] By contrast, departmental expenditure limits (DEL) overall are largely unchanged in nominal terms in this forecast and therefore lower in real terms relative to our March forecast. Maintaining the real value of DEL spending as in our March forecast would mean spending would need to be £20.5 billion higher by 2028-29. ^[82] In this sense, all of the headroom against the fiscal mandate at this event can be attributed to the erosion in the real value of departmental spending.

5.8 A large part of the net improvement in headroom against the fiscal targets since March can also be attributed to the rolling deadline for those targets. On a post-measures basis, the headroom against the debt falling target in 2027-28 is £1.6 billion in our latest forecast, a £4.9 billion deterioration on the headroom in March. So had the target date not rolled forward a year, the Chancellor would have only met his debt falling target by the narrowest of margins. Headroom against the supplementary borrowing target in 2027-28 remains relatively large at £44.1 billion.

Headroom against successive fiscal mandates

5.9 The £13.0 billion of headroom against the fiscal mandate in this forecast, while representing an increase on the November 2022 and March 2023 forecasts, is still well below the average headroom of previous Chancellors. This is also in the context of elevated uncertainty around our central forecast, particularly in relation to the outlook for inflation and productivity, discussed later in this chapter, the current strength in tax receipts as highlighted in Box 4.1, and pressures on departmental spending highlighted in Box 4.3.

5.10 Since the OBR was established in 2010, governments have had six different fiscal mandates. Chart 5.2 shows the headrooms against those mandates in successive forecasts. In most cases, Chancellors have chosen to aim for headroom of between £15 billion and £30 billion with an average of £29.7 billion (expressed in terms of today’s GDP). In several cases where rules were on course to be broken, such as in the wake of the coronavirus pandemic, new and less stringent ones were adopted. Most recently, in November 2022, the horizon for getting debt falling was extended from three to five years, which turned a prospective miss of £11.4 billion into headroom of £9.2 billion at the time.

Chart 5.2: Successive forecasts for headrooms 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. In this Autumn Statement, the Government has made a fiscally neutral adjustment to the welfare cap to reflect a recosting of the amount of spending shifted from inside to outside the cap due to the increase in the universal credit (UC) administrative earnings threshold to 18 hours. This recosting moves £0.7 billion of spending from inside to outside the cap and correspondingly reduces the level of the cap by £0.7 billion.

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 £8.6 billion in our latest forecast (Table 5.2). Spending subject to the welfare cap has been revised up by £6.2 billion relative to our March forecast. Excluding a £2.6 billion increase for higher CPI uprating and a small change of £0.2 billion to the Scottish welfare block adjustment, which are removed from the calculation of performance against the cap, the increase in the prospective breach of the cap primarily reflects:

• higher in-cap UC (£1.6 billion) in light of higher-than-expected caseloads and longer-than-expected benefit durations;
• the policy to restore the local housing allowance to the 30th percentile of local rents (increasing in-cap housing benefit and UC spending by £1.0 billion); and
• higher forecasts for housing benefit (£0.7 billion), and employment and support allowance spending (£0.4 billion).

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.^[83]

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.
• 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 3.5 per cent of GDP and 8.7 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. Thereafter, net interest costs decrease at a historically fast pace as RPI inflation falls. As a share of GDP, net interest costs remain above their pre-2007 level but compared to revenues are slightly below, reflecting the higher level of receipts in our latest forecast.

Table 5.3: Dashboard of balance sheet and fiscal affordability indicators

Box 5.1: The fiscal challenge of weaker growth and higher interest rates

Interest rates have risen significantly over the past two years and since our previous forecast in March. When closing the current forecast, market expectations for Bank Rate were 1.1 percentage points above the 4.3 per cent peak in mid-2023 assumed in our March 2023 forecast, while 10-year gilts had risen to a 15-year high of 4.5 per cent, 1.2 percentage points up from our March 2023 forecast. At the same time, our forecast for medium-term growth in nominal terms is 0.1 percentage points lower at the forecast horizon compared with March.

Since interest rates started rising in early 2021, the UK has seen a sharper rise in the costs of servicing its debt than other G7 economies. In 2022, the UK spent around 3.3 per cent of GDP on net interest payments compared with the G7 average of 1.7 per cent, with only Italy (which has a significantly higher debt stock) spending more as a share of its economy. And as a share of total receipts, the UK spent more on servicing its debt than any other G7 economy, spending 8.6 per cent of receipts compared to the G7 of average of 4.1 per cent in 2022.

Chart A: Net interest payments in G7 countries

As our 2023 Fiscal Risks and Sustainability report explored, the sharp rise in UK debt servicing costs has largely reflected the UK government’s relatively high share of index-linked gilts and the shortening of the average maturity of its net liabilities due to the Asset Purchase Facility. Looking forward, as RPI growth moderates and the APF unwinds, the pressures on the UK’s fiscal position are likely to reflect a more conventional interplay between the UK’s relatively weak medium-term economic growth prospects and relatively high levels of long-term interest rates.

The UK’s more challenging debt dynamics are illustrated in Chart B. UK 10-year government bond yields currently stand at 4.6 per cent – the second highest in the G7 behind the US and 1.0 percentage point higher than the G7 average. But the UK’s medium-term GDP growth is expected to be broadly similar to other G7 advanced economies. In its latest World Economic Outlook, the IMF forecasts medium-term nominal GDP growth to be 3.5 per cent, only slightly higher than the G7 advanced economy average of 3.2 per cent.

Chart B: Interest rates and forecast GDP growth in G7 countries

It is this difference between interest rates and nominal GDP growth (known as the growth-corrected interest rate or ‘r-g’) that is a key determinant for changes in debt as a share of GDP.^a Chart C shows the growth-corrected interest rate from across the G7 between 2007 and 2024. For most of the last 10 years, the UK has benefitted from a negative growth-corrected interest rate and a relatively more favourable dynamic compared to other advanced economies. But looking forward, these favourable dynamics are expected to deteriorate as the UK’s growth-corrected interest rate becomes positive. The OECD short-term forecasts imply a growth-corrected interest rate for the UK of 0.5 per cent in 2024, 0.9 percentage points higher than the G7 advanced economy average of -0.4 per cent.

Chart C: Growth-corrected interest rate across the G7 countries

In our latest forecast, the growth-corrected interest rate is forecast to reach 1.4 per cent in the medium term – a 1.0 percentage point deterioration on March.^b All else equal, this means that a larger primary surplus (the deficit excluding net interest payments) is required to reduce debt relative to GDP. Indeed, the primary surplus in our latest forecast stands at 2.2 per cent of GDP in 2028-29 – the highest since the OBR was established – but debt is only falling by 0.4 per cent of GDP in that year. It would therefore take only very small movements in either interest rates (up) or growth (down) to mean debt no longer falls as a share of GDP in the medium term.

a) Throughout this box, we use 10-year nominal bond yields when referring to long-term interest rates and for the calculation of the growth-corrected interest rate.

b) This is based on comparing the growth-corrected interest rate in the last year of the March forecast (2027-28) to the last year of this forecast (2028-29).

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 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.

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^[84]) 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 inflation and productivity.

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 5.3 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 56 per cent chance of underlying debt falling as a share of GDP in 2028-29 and 78 per cent chance of PSNB being less than 3 per cent of GDP in 2028-29. Relative to our March forecast, where these rules applied to 2027-28, the chance of debt falling is slightly higher (up from 52 per cent) as is the chance of meeting the borrowing target (up from 68 per cent).

Chart 5.3: Fan charts for PSND (excluding Bank of England) and PSNB

Sensitivities

5.20 Our sensitivity analysis tests 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, inflation, and departmental spending.
• The sensitivity of PSNB to changes in the effective tax rate, the effective interest rate, the level of GDP, inflation, and departmental spending.

The change in the debt-to-GDP ratio

5.21 We use our fiscal ready-reckoners to calibrate several possible adverse surprises relative to our central forecast that would be sufficient to negate the 0.4 per cent of GDP year-on-year fall in debt (excluding the Bank of England) in 2028-29.^[85]
It could fall to zero if:

• The effective tax rate was 0.4 percentage points lower. This would be equivalent to the tax-to-GDP ratio rising by only 1.1 percentage points between 2022-23 and 2028-29 rather than rising by 1.5 percentage point in our central forecast.
• Effective interest rates on central government debt were 0.4 percentage points higher at 4.0 per cent. This is equivalent to just under one-tenth of the Bank of England’s 5 percentage point hike in the Bank Rate over the latest hiking cycle.
• Nominal GDP was 0.4 per cent lower, via its effects across multiple tax bases. This could result from average GDP growth being about 0.1 percentage points a year lower than in our central forecast.
• Departmental spending was 2.2 per cent or £13.0 billion higher. Restoring DELs to their March levels in real terms in 2028-29 would cost £20.5 billion and therefore more than eliminate the headroom to debt falling in that year.

Public sector net borrowing

5.22 Our central forecast is for PSNB to be 1.1 per cent of GDP in 2028-29, giving headroom of £61.5 billion against the target for PSNB to not exceed 3 per cent of GDP. This is large relative to the headroom the Chancellor has against his debt falling target, but it could still fall to zero if:

• The effective tax rate was 1.9 percentage points lower, equivalent to the tax-to-GDP ratio falling by 0.4 percentage points between 2022-23 and 2028-29 rather than rising by 1.5 percentage points in our central forecast.
• Effective interest rates on central government debt were 1.8 percentage points higher at 5.5 per cent, broadly equivalent to their level in the first half of the 2000s.
• Nominal GDP was 3.8 per cent lower, via its effects across multiple tax bases. This could result from average GDP growth being about 0.7 percentage points a year lower in our central forecast. The overall shortfall would be equivalent to about two-thirds of the peak-to-trough fall in GDP in the recession caused by the global financial crisis.
• Departmental spending was £61.5 billion (10.6 per cent) higher. This would be the equivalent to departmental spending increasing to 20.0 per cent of GDP in 2028-29 instead of falling to 18.1 per cent of GDP in our central forecast.

Scenarios

5.23 Our latest forecast features higher inflation, higher interest rates, and slower potential output growth than we expected in March. This has proved to be favourable to the public finances, largely because of the source of the inflation (domestic rather than external) and the Government’s decision to allow higher inflation to erode the real value of departmental spending. But different configurations of inflation and growth are also possible, and could be more or less advantageous fiscally. To illustrate this, these scenarios explore the fiscal implications of alternative paths for: (i) the level and driver of inflation; and (ii) the underlying rate of productivity growth.

Imported versus domestic inflation

5.24 One of the biggest changes in this forecast compared to our March outlook is to the level and composition of inflation. In this forecast, annual inflation averages 3.2 per cent from 2023 to 2027 compared with 1.8 per cent in our March forecast. We previously expected above-target inflation to be mainly driven by higher imported energy and other goods prices, but we now forecast more domestically driven inflation. We have considered before how this kind of change might affect the fiscal position, in Box 2.6 of our October 2021 Economic and fiscal outlook.

5.25 To illustrate the difference between imported and domestically generated inflation for the public finances, we present scenarios for a more persistent version of each. In both scenarios, CPI inflation is 1.5 percentage points higher by the middle of next year and then takes three years longer to return to our baseline forecast. This means inflation remains above 6 per cent until the second quarter of 2024, and above 4 per cent until the end of next year. In response, the Bank of England is assumed to raise interest rates further to a peak of 7 per cent in early 2024, 1.6 percentage points above our baseline peak.^[86]
The difference between the two inflation scenarios Chart 5.4 is that:

• in the pure imported inflation scenario (due, for example, to higher global energy prices) the shock pushes up consumer prices but leaves the price of UK output unchanged – so nominal GDP, wages, and profits are the same as in the baseline.
• by contrast, the pure domestic inflation scenario (due to a tighter labour market, for example) sees higher consumer prices reflected fully in higher nominal GDP.

Chart 5.4: The size of the inflation shocks and how they pass into nominal GDP

Stronger and weaker potential productivity

5.26 Inflation is obviously not the only factor that could change our nominal GDP profile – changes in real output growth can also be a driver. Productivity growth – the efficiency with which labour is used to produce output – tends to be the most unpredictable element of overall economic growth. Since 2010, over-optimism on potential productivity has been a major source of the OBR’s past economic and fiscal forecast errors.^[87]
Looking forward, we expect a slight rise in productivity growth from recent subdued rates to an average annual rate of around 1 per cent over the next five years. But to illustrate the sensitivity of the public finances to future productivity growth, as shown in Chart 5.5, we also consider:

• an upside scenario in which annual productivity growth is ½ a percentage point higher at around 1½ per cent, which would see productivity growth returning toward its pre-financial crisis average, perhaps due to an acceleration in the pace of technological change, such as the widespread adoption of artificial intelligence. This raises 2028-29 nominal GDP by roughly the same amount as in the domestic inflation scenario.
• a downside scenario in which annual productivity growth is ½ a percentage point lower at around ½ a per cent, which would be close to the average rate of productivity growth in the decade following the financial crisis.

Chart 5.5: Potential productivity growth

The fiscal consequences of alternative inflation and productivity scenarios

5.27 These different potential shocks to the economic outlook have very different consequences for the medium term fiscal position. Starting with the inflation scenarios as shown by the left-hand panel of Chart 5.6:

• In both the imported and domestic inflation scenarios, debt interest spending is higher due to both higher interest rates and RPI (which raises the cost of uprating index-linked gilts). Spending on welfare and pensions which are indexed to CPI inflation, is also higher.
• In the imported inflation shock, there is a significant deterioration in the public finances. Relative to the baseline, borrowing is £23.2 billion higher by 2028-29, as shown in the green line in Chart 5.6. That is because the imported inflation does not lift domestic wages, profits, or other major nominal tax bases.
• In the domestic inflation shock, the net fiscal position is either unchanged or significantly improved depending on decisions taken about the path of departmental spending. This is because domestically generated inflation boosts all major nominal tax bases and increases the effective tax rate (via more ‘fiscal drag’), which more than offsets the rise in debt interest and welfare costs. If departmental budgets are adjusted to preserve their real spending power, then the net impact on borrowing is close to zero over the forecast, as shown in the dotted yellow line. If, by contrast, departmental budgets remain unchanged in cash terms, as they largely are in this forecast, the erosion in their real value delivers a £15.8 billion reduction in borrowing relative to the baseline by 2028-29, as shown in the solid yellow line.

5.28 Alternative scenarios for potential productivity growth can have similarly material fiscal implications as shown by the right-hand panel of Chart 5.6:

• In the upside productivity scenario, receipts are £39.9 billion higher than in our central forecast by 2028-29, while spending is £6.2 billion lower. The lower spending is largely explained by the reduction in debt interest payments, resulting from the lower central government net cash requirement. The net result is that borrowing is £46.0 billion lower by the end of the forecast period relative to our central scenario (orange line in Chart 5.6).
• In the downside productivity scenario, the fiscal impacts are broadly symmetric, with lower receipts and higher debt interest spending raising borrowing by £42.2 billion by 2028-29 relative to our baseline.

Chart 5.6: Borrowing in the inflation and productivity scenarios

5.29 Looking at what these different scenarios implies for the dynamics of public debt, Chart 5.7 shows the trajectory for debt as a share of GDP under each, and shows that:

• Debt as a share of GDP falls fastest in the higher productivity scenario. Headroom against debt falling in the fifth year of the forecast increases to £72.0 billion, compared to £13.0 billion in our central forecast. The improved headroom reflects a lower primary deficit and a permanent improvement in the growth-corrected interest rate (the effective interest rate on the debt stock minus nominal GDP growth).
• The two inflation scenarios have different impacts on the trajectory of debt. In the imported inflation scenario, headroom against debt falling in the fifth year of the forecast is missed by £1.5 billion, whereas in the domestic inflation scenario the headroom is met by £37.4 billion. However, if departmental spending is held constant in real terms in the domestic inflation scenario, headroom narrows to £23.6 billion.
• Debt as a share of GDP is rising fastest in the lower productivity scenario. With the primary deficit now wider and the growth-corrected interest rate permanently higher, the debt target is missed by £40.5 billion.

Chart 5.7: Public sector net debt as a share of GDP in the scenarios

5.30 These scenarios illustrate that while more domestically generated inflation can appear to help the fiscal position, the improvement depends on there being a greater squeeze in real government spending. By contrast an imported inflation shock leads to an unambiguous deterioration in the fiscal position. However, it is the slower productivity growth scenario that poses the greatest risk to fiscal sustainability, while the higher productivity scenario is the most fiscally beneficial. This illustrates that shocks to productivity – the key driver of long-run economic growth – are likely to have the most significant and persistent implications for fiscal dynamics.

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 March 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 March

Table A.3: Determinants of the fiscal forecast

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

Table A.5: Current receipts

Table A.6: Current receipts: changes since March

Table A.7: Total managed expenditure