Fiscal risks report

Office for
Budget
Responsibility

Fiscal risks report

July 2021

July 2021

CP 453
Office for Budget Responsibility
Fiscal risks report

Presented to Parliament by the

Exchequer Secretary to the Treasury
by Command of Her Majesty

July 20

CP 453
 Cro copyri t 2021

is pu licatio is lice se u er t e ter s of t e Ope o er e t

ice ce 3 0 e cept ere ot er ise state o ie t is lice ce isit
atio alarc i es o uk oc ope - o er e t-lice ce ersio 3

ere e a e i e ti e a y t ir party copyri t i for atio you ill

ee to o tai per issio fro t e copyri t ol ers co cer e

is pu licatio is a aila le at o uk official- ocu e ts

y e uiries re ar i t is pu licatio s oul e se t to us at

o r e uiries o r uk

ISBN 978-1-5286-2649-1

CCS0521649260 07 21

Pri te o paper co tai i 75 recycle fi re co te t i i u

Pri te i t e y t e PS roup o e alf of t e Co troller of

er a esty s Statio ery Office

 Contents

Foreword ...................................................................................... 1

Executive summary ........................................................................ 3

Chapter 1 Introduction ................................................................................ 23

Chapter 2 Coronavirus pandemic ................................................................ 37

Chapter 3 Climate change .......................................................................... 83

Chapter 4 Cost of public debt .................................................................... 153

Chapter 5 Update on other fiscal risks ........................................................ 207

Index of charts and tables .......................................................... 233

Foreword

The Office for Budget Responsibility (OBR) was established in 2010 to examine and report on the
sustainability of the public finances. A central feature of our efforts to meet that remit has been
finding better ways to capture and communicate economic and fiscal risks. Ever since our first
Economic and fiscal outlook (EFO) in 2010, we have emphasised the degree of uncertainty around
our central economic and fiscal forecasts by showing probabilistic ranges (‘fan charts’) for the key
economic and fiscal aggregates based on historical forecast errors. Our biannual EFOs also feature
alternative scenarios and sensitivity analysis to illustrate the implications of changing key forecast
judgements. The long-term projections in our Fiscal sustainability reports (FSRs) include sensitivity
analysis to changes in key demographic, macroeconomic, and other assumptions. And we have
relied on scenario analysis more than ever over the past year to illustrate the huge uncertainties
surrounding the path of the coronavirus pandemic and the public health and fiscal policy response.

In the October 2015 update to the Charter for Budget Responsibility, Parliament required us to
produce a report on fiscal risks at least once every two years, and for the Government to respond
formally to each report within a year. We produced our first Fiscal risks report (FRR) in July 2017 and
the Government responded in Managing fiscal risks in July 2018. Our 2019 FRR reflected on the
Government’s response and extended the dialogue. We continue the series of exchanges in this
report, taking account of the Government’s understandably limited written response to our previous
report, and develop the discussion in the light of the crystallisation of one of the largest risks in
centuries, the coronavirus pandemic, and the unprecedented array of policy measures introduced to
mitigate its economic and fiscal impact.

Our previous two FRRs took an encyclopaedic approach, aiming to provide a full account of the –
mostly adverse – risks to the public finances. One of the main conclusions of both was that major
shocks to the public finances are inevitable, if unpredictable, so governments need to recognise that
they are very likely to have to confront them at some point. But while those two editions identified
more than 90 different fiscal risks, neither considered in any detail the potential economic and fiscal
consequences of a global pandemic. However, the fiscal stress test included in our first FRR did
presage the roughly 30 per cent of GDP rise in government debt resulting from the pandemic, albeit
as a result of a different combination of shocks. This underscores the value to fiscal forecasters and
policymakers of exploring, and trying to communicate to the public, the nature and scale of
potential shocks to the public finances, even if their precise nature, timing, and magnitude is
uncertain. Exploring the consequences of a global pandemic – which sat atop the Government’s
2015 National Risk Register – would of course have been more valuable still.

This FRR has been prepared in the wake of the largest fiscal risk to have crystallised in peacetime –
the coronavirus pandemic – the economic and fiscal consequences of which continue to be felt. It
therefore focusses on what we can learn from this experience to enhance our understanding, and
inform the Government’s management, of other potentially catastrophic or ‘tail risks’ facing the UK
and other countries around the world. So, this FRR departs from the encyclopaedic approach of past

1 Fiscal risks report

Foreword

reports and shifts focus onto three sources of potentially very large fiscal risks: the coronavirus
pandemic, climate change, and the cost of public debt. It also provides updates on significant
developments regarding the specific risks highlighted in previous FRRs.

The analysis and conclusions presented in this document represent the collective view of the three
independent members of the OBR’s Budget Responsibility Committee. We take full responsibility for
the judgements that underpin them. We have been hugely supported by the staff of the OBR, to
whom we are as usual enormously grateful, particularly in the wake of a very demanding year.

We have also drawn on the help and expertise of officials across numerous government
departments and agencies, including HM Treasury, the Bank of England, Climate Change
Committee, Debt Management Office, Department for Business, Energy and Industrial Strategy, HM
Revenue and Customs, Joint Biosecurity Centre, National Infrastructure Commission and the
Scientific Pandemic Influenza Group on Modelling. We are very grateful for their insight.

In addition, we have benefitted from discussions with experts from outside government who have
spoken to us about our three main topics. In particular, we would like to thank colleagues at the
IMF, Jan Vlieghe of the Bank of England’s Monetary Policy Committee, Dimitri Zenghelis at the
Bennett Institute at Cambridge University, Kevin Daly at Goldman Sachs, Anita Charlesworth and
colleagues at the Health Foundation’s REAL centre, Paul Johnson and Carl Emmerson at the Institute
for Fiscal Studies, Andrew Scott at the London Business School, Tony Travers and Lukasz Rachel from
the London School of Economics, Bill Allen at the National Institute of Economic and Social
Research, Torsten Bell and colleagues at the Resolution Foundation, and Ian Mulheirn and Tim Lord
at the Tony Blair Institute for Global Change. Finally, we are particularly grateful to Sarah Breeden
and her team at the Bank of England for sharing insight and analysis on their climate change
scenarios, which we draw on in this report. We would also emphasise that despite the valuable
assistance received, all judgments and interpretation underpinning the analysis and conclusions of
the FRR are ours alone.

We provided the Chancellor of the Exchequer with a summary of our main conclusions on 25 June.
Given the importance of the report to the Treasury in managing fiscal risks, we have engaged with
officials there regularly throughout and requested their assistance in understanding developments
since our previous report in order to enrich our analysis. We provided an advance pre-release copy
on 2 July and a full and final copy 24 hours prior to publication, in line with pre-release access
arrangements set out in the Memorandum of Understanding between the Office for Budget
Responsibility, HM Treasury, Department for Work and Pensions and HM Revenue & Customs. At no
point in the process did we come under any pressure from Ministers, special advisers or officials to
alter any of our analysis or conclusions.

Richard Hughes Sir Charles Bean Andy King

The Budget Responsibility Committee

Fiscal risks report 2

Executive summary

1 Just two decades into this century, the UK has already experienced two ‘once in a century’
economic shocks – the 2008 financial crisis and the 2020 coronavirus pandemic. These two
shocks triggered the two largest post-war recessions, accounted for successive peacetime
government borrowing records, and added over £1 trillion (50 per cent of GDP) to public
debt – taking it above 100 per cent of GDP for the first time since 1960. While these shocks
have very different origins, impacts, and likely legacies, both offer a stark reminder of the
importance of understanding risk to effective fiscal forecasting and policymaking.

2 As we emerge from the largest peacetime economic and fiscal shock in three centuries, our
third Fiscal risks report (FRR) departs from the encyclopaedic approach of our previous two
and shifts focus onto three sources of potentially very large fiscal risks: the coronavirus
pandemic, climate change, and the cost of government debt. These three risks are very
different in nature, but nevertheless have some important features in common. There is a
high degree of uncertainty concerning both their timing and associated costs. They are
characterised by non-linearities or ‘snowball effects’ in which costs can escalate dramatically
from the point of crystallisation. And they are global in nature, with the potential for rapid
contagion across countries. Governments seeking to manage these threats must thus weigh
the known costs of early action to mitigate these risks against the uncertain costs of dealing
with the fallout when they crystallise. They must also weigh the limited but more deliverable
benefits of acting unilaterally against the greater but more elusive gains from acting globally.

Increasing fiscal exposure to catastrophic risks

3 Chapter 1 defines what we mean by fiscal risks, outlines our approach to analysing them,
and considers whether governments are becoming more exposed to potentially catastrophic
risks. Fiscal risks are factors that lead fiscal outcomes to deviate from forecasts over the
medium term or threaten fiscal sustainability over the long term. Our FRRs tend to focus on
adverse or ‘downside’ risks. This asymmetric approach reflects the tendency for shocks to the
public finances (especially large ones) to be bad rather than good, as well as governments’
tendency to spend, rather than save, unexpected windfalls, but to absorb, rather than offset,
unexpected costs. It is, of course, a central function of government to pool and manage risks
that cannot be borne solely by individuals or firms. During normal times, governments
provide varying degrees of ‘social insurance’ for the unemployed, the sick, and the old.
During crises, in which multiple risks tend to crystallise at once, governments can and do
take on a much broader range of costs by virtue of their role as ‘insurer of last resort’.

4 The arrival of two major economic shocks in quick succession need not constitute a trend,
but there are reasons to believe that advanced economies may be increasingly exposed to
large, and potentially catastrophic, risks. While the threat of armed conflict between states

3 Fiscal risks report

Executive summary

(especially nuclear powers) appears to have diminished in this century, the past twenty years
have seen an increase in the frequency, severity, and cost of other major risk events, from
extreme weather events to infectious disease outbreaks to cyberattacks (Chart 1). And
estimates from major insurers and others of the amount of global GDP at risk from these
and other potentially catastrophic risks have been rising steadily. This appears to reflect a
combination of the increased frequency and severity of some anthropogenic risks (such as
climate change and cyberattacks), growing numbers of people living and working in greater
proximity to the sources of those risks (such as floodplains and isolated ecosystems), and
deepening global interconnectedness (through travel, trade, finance, and the internet).

Chart 1: Incidence of major risk events

500
Climatological
450
Geophysical
400
Meteorological
350
Hydrological
Number of events

300

250

200

150

100

50

0
1980 1985 1990 1995 2000 2005 2010 2015
Source: The Emergency Events Database

5 As countries’ exposure to large, and potentially catastrophic, risks increases, so do the

associated risks to their public finances. This is because such risks are not only more
disruptive to the economies that generate governments’ revenues but also because they are
more likely to overwhelm private risk management and insurance mechanisms, prompting
governments to step in as insurer of last resort. This may be particularly true in an era when
economic shocks are more severe, financial institutions and firms are more leveraged, and
monetary policy is more constrained. So knowing where risks reside and how they spread
across economies and onto government balance sheets is central to understanding the
evolution of the public finances over the past two decades and the potential threats to fiscal
sustainability over the rest of this century.

Coronavirus pandemic
6 Chapter 2 therefore looks back at the fiscal impact of the coronavirus pandemic over the
past year and ahead to its potential legacy for the public finances over the medium and long
term. It would be premature to draw definitive conclusions about the economic and fiscal

Fiscal risks report 4

 Executive summary

consequences of the pandemic while the virus continues to circulate and mutate, economic
activity remains subject to public health restrictions, and extensive fiscal support remains in
place. But it is nonetheless instructive to look at the UK’s experience to date, in both historical
and international context, as a case study in potentially catastrophic fiscal risk.

Economic impact
7 The pandemic brought about the largest and most synchronised peacetime shocks the world
economy has faced since the Great Depression of the 1930s. Global output fell by 3.3 per
cent in 2020, far greater than the 0.1 per cent fall seen during the global financial crisis in
2009. Regardless of how successfully they insulated themselves from the virus itself, few
countries escaped its economic consequences. Almost 90 per cent of economies suffered a
decline in output last year, including every advanced economy except Taiwan and Ireland
(Chart 2). While the pandemic is not yet over, countries that were able to contain the spread
of the virus early have so far typically experienced shorter and shallower downturns and
faster recoveries, on average returning to pre-pandemic levels of activity at the start of 2021.

Chart 2: Fall in real GDP in 2020 in advanced economies

4

2
2020 change in GDP (per cent)

-2

-4

-6

-8

-10

-12

Taiwan
Australia

Korea
United Kingdom

Belgium

Ireland
Luxembourg
Finland

Norway
Slovak Republic

Lithuania
Iceland

Switzerland
Canada

Sweden
Portugal

Austria

Germany
France

Slovenia

Latvia
Cyprus
Hong Kong

Denmark
Greece

Singapore
Spain

Italy

Japan

United States

New Zealand

Estonia
Malta

Czech Republic

Israel
Netherlands

Note: Based on IMF definition of advanced economies.

Source: IMF, ONS

8 The UK suffered one of the deepest recessions among advanced economies, with UK GDP
falling by 10 per cent in 2020 as a whole, twice the advanced economy average. The
relative severity of the downturn in the UK last year compared with other advanced
economies, which is only partly reduced when looking at alternative measures that allow for
cross-country methodological differences in measuring real output,1 is likely to be a
consequence of our being relatively hard hit by the virus itself (suffering among the highest

1
Box 2.4 of our March Economic and fiscal outlook looked in detail at international comparisons of the economic impact of the pandemic
up to the third quarter of 2020. On the current vintages of data, the UK experienced a larger fall in output in 2020 than most other major
advanced economies, even after adjusting for differences in the measurement of government output.

5 Fiscal risks report

Executive summary

rates of infections, hospitalisations, and deaths in 2020), spending more time under stricter
public health restrictions (second only to Italy), and being more economically vulnerable to
the pandemic due to our large share of social consumption in output (second only to Spain).

9 However, the economy has also proved surprisingly adaptable and resilient to the
coronavirus shock. The relationship between the stringency of public health restrictions and
levels of economic output weakened significantly over the course of the pandemic. And
economic activity rebounded quickly once those restrictions were eased. So while output after
the 2008 financial crisis did not return to its pre-crisis level for more than 4½ years, our
latest forecast assumes it will regain its pre-pandemic level by the middle of next year, just
over two years since coronavirus arrived in the UK. This economic resilience is likely to reflect
the lack of any overheating of the economy going into the pandemic, coupled with the
unprecedented amount of fiscal support provided to all parts of the economy which helped
keep firms liquid and solvent and employees attached to their employers.

Fiscal impact
10 Reflecting its outsized impact on the UK economy, the pandemic also imparted an
extraordinary shock to the UK public finances. At the time of our March forecast we expected
it to push government borrowing to a peak of 16.9 per cent of GDP in 2020-21, the highest
since 1944-45, and for public sector net debt to reach a peak of 108.6 per cent of GDP in
2023-24, its highest level since 1958-59. The UK saw the fourth largest increase in
government borrowing among 35 advanced economies (after Canada, Norway and
Singapore) in 2020. And in contrast to what happened during the financial crisis, the bulk of
the increase in cash borrowing was the result of discretionary increases in government
spending rather than the impact of the crisis on government receipts.

11 The UK’s fiscal policy response to the pandemic was large by both historical and
international standards (Chart 3). The UK’s coronavirus rescue package cost 16.2 per cent of
GDP over 2020-21 and 2021-22, almost ten times that provided during the financial crisis
in 2008-09 and 2009-10. This was the third largest among 35 advanced economies after
the United States and New Zealand, and was also more heavily skewed toward spending on
healthcare, making up a third of total pandemic-related spending in the UK versus less than
15 per cent on average across the advanced economies. The relative size of the UK’s fiscal
policy response is likely to be a consequence of several factors including the pandemic’s
outsized impacts on the economy and health services, the fact that the UK entered the
pandemic with relatively little spare capacity in the health service, and a pre-pandemic
system of working-age welfare support that replaced less of the incomes of those losing
hours or falling out of work. This extent of the overall rescue package not only protected
household and firm incomes, but also tax revenues, which fell much less than one would
expect given the dramatic fall in GDP.

Fiscal risks report 6

 Executive summary

Chart 3: Fiscal rescue packages across major economies

30
Other
Public works
25
Firms
Employment
20
Households
Per cent of GDP

Healthcare
15

10

0
Spain France Italy Germany Canada Japan United United Average
Kingdom States
Note: These figures and chart average come from IMF estimates of measures between January 2020 and April 2021 in 10 economies
for which detailed data are available.
Source: IMF, OBR

12 Pandemic rescue packages in the UK and across the advanced economies made
unprecedented use of unconventional fiscal instruments, some of which carry risks for public
finances well into the future. Advanced economy governments extended about as much in
loans, guarantees, and other forms of quasi-fiscal support in 2020 as they did in
conventional tax reliefs and subsidies. The UK made relatively active use of these
instruments, guaranteeing 16 per cent of GDP in loans and about half of all lending to small
and medium-sized businesses in 2020-21. While their ultimate cost is highly uncertain, our
latest forecast assumes that around one-third of the total value of these loans will end up
being covered by the taxpayer.

Medium-term fiscal legacy

13 A relatively rapid economic recovery (supported by the UK’s early rollout and high take-up of
vaccines) and the withdrawal of pandemic-related fiscal support should lead to a sharp fall
in borrowing this year. But there are significant risks to the Government’s medium-term fiscal
plans from the legacy of direct funding pressures that the pandemic may leave behind for
public services. Departmental spending plans make no provision for virus-related spending
beyond this financial year. Instead, spending totals from 2022-23 onwards were cut by
£14½ billion a year in Spending Review 2020 and the 2021 March Budget relative to the
sustained rises in departmental spending planned pre-pandemic. At the same time, overall
public spending is still forecast to be higher as a share of GDP in the medium term than it
was pre-pandemic.

7 Fiscal risks report

Executive summary

14 The extent to which any additional spending to meet pandemic-legacy pressures leads to
higher departmental resource spending (RDEL) overall would depend on choices made at
future Spending Reviews, starting with the next one this Autumn. The Government might, for
example, keep total spending over the coming years unchanged and choose to allocate less
than it otherwise would have done to its pre-pandemic priorities given the changed
circumstances. Or it could choose to increase total spending, which would either require
further tax rises or put at risk the Chancellor’s aim of balancing the current budget and
getting debt falling by the middle of the decade.

15 In any case, these potential unfunded legacy costs of the pandemic represent a material risk
to the public spending outlook. Considering just selected pressures in three major spending
areas, the Government could face spending pressures of around £10 billion a year on
average in the next three years. These include:

Health. Pressures on health budgets could be around £7 billion a year from the
potential need to pay for: standing test and trace and revaccination programmes; the
consequences of the pandemic for individuals’ physical and mental health; additional
spare capacity to cope with possible future outbreaks; and the pandemic-related
backlog of treatments.

Education. Schools may require around £1¼ billion a year to enable pupils to catch
up on the estimated two to three months of education that they have lost on average
during the pandemic, in addition to the £1.4 billion that has been committed since the
Budget, with the intention of reviewing the case for further funding in the Spending
Review.

Transport. Around £2 billion a year may be needed to fill a 10 to 25 per cent hole in
the fare revenues of the new Great British Railways and Transport for London (TfL) if
passenger numbers do not return to pre-pandemic levels. The Government has
already provided £12.8 billion of direct support to the railways and TfL in 2020-21.
However, as of June 2021, passenger numbers on national rail and the London
Underground were still down a half on pre-pandemic levels.

Fiscal risks report 8

 Executive summary

Chart 4: Illustrative estimates for selected pandemic-related pressures on

departmental resource spending
14
Transport
Education
Health: Test & Trace, and vaccinations
12
Health: Pandemic legacy pressures
Health: NHS backlog
10

8
£ billion

0
2022-23 2023-24 2024-25
Note: Assumptions and sources underpinning estimates of individual pressures can be found in Chapter 2.
Source: OBR

Long-term economic legacy

16 Beyond these direct medium-term pressures, the longer-term fiscal risks associated with the
pandemic will depend on its lasting impact on potential GDP. As set out in our last two
Economic and fiscal outlooks, we have so far assumed a ‘scarring’ effect (defined as the
shortfall of potential output relative to the pre-pandemic trajectory at the five-year forecast
horizon) of 3 per cent in our central forecast. This scarring results from lower investment,
lower labour supply, and lower total factor productivity in roughly equal proportions.

17 Evidence to date on the potential degree of scarring has been mixed. There has been some
upside news on the paths of both GDP and investment, but against that there has been
downside news on net outward migration over the past year. And there remains considerable
uncertainty regarding the future size of the workforce due to continuing lower net inward
migration and the effect of the pandemic on participation and hours. Some forecasters,
including the Bank of England and IMF, have lowered their estimates of scarring from the
pandemic in the light of recent developments. We should learn more about the effects of the
pandemic in the coming months as remaining public health restrictions are lifted, the
furlough and other business support schemes wind down, and borders reopen. But it is the
medium and longer-term outlook for GDP – which will reflect a combination of pandemic
effects, Brexit effects, and assumptions about underlying potential output growth – that
matters for the sustainability of the public finances.

9 Fiscal risks report

Executive summary

Climate change
18 Chapter 3 looks ahead to the fiscal risks presented by climate change, and the economic
and fiscal implications (both positive and negative) of alternative paths to meeting the
Government’s legislated goal to reduce net greenhouse gas emissions to zero by 2050. The
fiscal risks from climate change can be split into those stemming from global warming itself
(‘physical risks’) and those relating to the move to a low-carbon economy, including the
policies to achieve that (‘transition risks’). In unmitigated climate change scenarios, the
physical risks dominate, whereas the more that is done to mitigate global warming by
reducing emissions, the more important transition risks become.

19 Climate change results from several market failures – most importantly that the costs of
emissions to current and future generations are not borne by those who produce them today.
This can be addressed by applying an appropriate price on carbon (for example via a tax or
an emissions trading scheme (ETS)). But there are many other policy challenges to overcome,
so the path to net zero can be expected to involve many policy levers on top of carbon taxes
and ETSs, including bans and other regulations, and public subsidies and investment. These
will all have economic and fiscal implications of one sort or another – either directly (via
taxes and spending) or indirectly (via wider economic outcomes).

The transition to net zero

20 By international standards, the UK has made good progress in reducing emissions over the
past 30 years, but there are greater challenges ahead. As of 2019, UK greenhouse gas
emissions were down 44 per cent relative to 1990. In particular, the source of power
generation with the highest emissions – coal – has disappeared from the energy mix thanks
to concerted policy efforts. Getting the rest of the way to net zero by 2050 will require us to
find ways of overcoming both the technological obstacles to delivering cost-effective carbon
removals at scale, and the delivery challenges associated with upgrading insulation and
installing low-carbon heating systems in more than 28 million homes.

21 Between now and 2050 the fiscal costs of getting to net zero in the UK could be significant,
but they are not exceptional. In net terms they will entail any direct public spending on the
cost of transition, receipts lost from existing emissions-related taxes (especially fuel duty),
receipts gained from taxing carbon more heavily, and the indirect effects of different paths
for the economy on the public finances. To construct paths for each of these, we draw on
scenarios produced by the Climate Change Committee (CCC) for whole economy costs and
savings from decarbonisation, and by the Bank of England for the price of carbon necessary
to achieve net zero and its economic implications. In the Bank’s ‘early action’ scenario
(which we use as our reference scenario), the imposition of a higher and steadily rising
carbon price weighs on economic activity, with GDP settling 1.4 per cent below its (purely
hypothetical) counterfactual path in which there are no additional climate-related headwinds.

Fiscal risks report 10

 Executive summary

An early action scenario

22 The fiscal impact of achieving net zero in the early action scenario adds 21 per cent of GDP
to public sector net debt in 2050-51 (£469 billion in today’s terms). That is somewhat
smaller than the addition to net debt as a result of the pandemic. It reflects:

Net zero public spending. The CCC puts the cumulative investment cost for the whole
economy between now and 2050, plus the operating costs of emissions removals, at
£1.4 trillion in 2019 prices. The Government has not said how much of that cost it
expects to bear. Our scenario assumes that public spending meets around a quarter of
it. When combined with savings from more energy-efficient buildings and vehicles, the
net cost to the state is £344 billion in real terms. But spread across three decades, this
represents an average of just 0.4 per cent of GDP a year.

Net zero receipts losses. Fully electric vehicles pay no fuel duty and are exempt from
vehicle excise duty (VED), so receipts from both fall almost to zero by 2050-51. Some
smaller tax bases (air passenger duty, landfill tax, the plastic packaging tax) are hit
too. Overall, receipts worth 1.6 per cent of GDP are lost in 2050-51, with fuel duty
accounting for 76 per cent and VED for 18 per cent.

Carbon tax revenues. Our scenario assumes all emissions are taxed, and more
heavily, from 2026-27 onwards (which could be achieved by extending the UK ETS or
imposing a uniform carbon tax in its place). Based on elements of the Bank and CCC
scenarios, the tax rate starts at £101 per tonne (in real terms) and rises steadily to
reach £187 per tonne in 2050-51. On this basis, additional carbon tax revenues raise
1.8 per cent of GDP in 2026-27, after which revenues decline steadily to 0.5 per cent
of GDP in 2050-51 as falling emissions more than outweigh the effect of the rising tax
rate. Towards the end of this time frame revenues are very uncertain, with an
increasingly narrow tax base and an increasingly high tax rate, meaning even small
differences in the pace of emissions abatement would have large revenue impacts.

Indirect fiscal consequences. We assume that public services and non-climate-related

investment are maintained in real terms in the face of modestly lower GDP. That raises
public spending as a share of GDP by 0.3 percentage points on average.

Debt interest costs. The higher path for debt increases debt interest spending by
increasing amounts, particularly towards the end of the period when fuel duty losses
are greatest. Additional debt interest reaches 0.7 per cent of GDP in 2050-51.

11 Fiscal risks report

Executive summary

Chart 5: Early action scenario: impact on public sector net debt

40
Debt interest spending
Indirect effects
30
Net zero public spending
Net zero receipts losses
20
Denominator effect
Per cent of GDP

Carbon tax revenues

10
Overall effect

-10

-20
2020-21 2023-24 2026-27 2029-30 2032-33 2035-36 2038-39 2041-42 2044-45 2047-48 2050-51
Source: OBR

Alternative scenarios and sensitivities

23 The economic and fiscal consequences of the transition to net zero are subject to many
sources of uncertainty. We therefore also consider alternative scenarios and sensitivities
(Chart 6). These capture the implications of:

Unmitigated climate change. The fiscal risks from extreme, unmitigated climate
scenarios cannot be quantified with any precision. But to give a sense of the potential
orders of magnitude, we produce an illustrative path for debt if average UK
temperatures were to rise by around 4°C by the end of this century, relative to the
average over the 20 years to 2000. (We use this longer horizon because the UK is
relatively insulated from climate change in the next few decades.) This entails greater
economic and fiscal costs to adapt to higher temperatures, but more importantly it is
assumed to result in progressively more frequent and more costly shocks to the public
finances than have historically been the case, reflecting both extreme weather events at
home and the spillovers from even greater damages in hotter countries. Relative to a
baseline that incorporates only the historical frequency and cost of such shocks, debt
ratchets up more sharply to reach 289 per cent of GDP by the end of the century.

Delayed action. To test assumptions about the timing and smoothness of action to
deliver the transition by 2050, we use the Bank’s ‘late action’ scenario. Decisive steps
to cut emissions globally and in the UK are delayed until 2030, then introduced
abruptly to deliver the necessary reductions in the shorter period left to the target date,
causing economic disruption and the premature scrapping of some capital. The main
differences to the early action scenario are that GDP settles around 3 per cent lower

Fiscal risks report 12

 Executive summary

still, while direct public spending costs increase by around a half. Overall, debt in
2050-51 is 23 per cent of GDP higher than in the early action scenario.

Uncertain consequences for productivity. The costs of low-carbon technologies could

fall further and faster than assumed, so an optimistic variant assumes decarbonisation
actually boosts productivity (by 0.1 percentage points a year). But equally, the costs
associated with such a major structural change over a sustained period could easily be
greater than assumed – for example, if a technology that has been heavily invested in
proves unsuccessful – so a pessimistic variant assumes it weighs more heavily on
productivity (by the same margin). In the high productivity variant, the larger economy,
and the lower path for public spending as a share of GDP that results, lowers debt by
11 per cent of GDP relative to the early action scenario by 2050-51. The results are
approximately symmetrical in the low productivity variant.

High versus low public sector share of net zero investment. Our reference early action
scenario assumes the state pays around a quarter of the total direct cost of the
transition, but the public sector’s share of investment in decarbonisation could vary
greatly. At the minimum, it will need to meet the costs associated with public sector
buildings and vehicles. At the higher end of the spectrum, it could deliver a much
greater share of net zero infrastructure, for example to overcome inertia in areas like
the domestic heating transition where progress to date has been slower. Our low
spending variant, in which the state bears around an eighth of the whole economy
costs, results in debt in 2050-51 being 5.2 per cent of GDP below the early action
scenario. In the high spending variant, in which the state bears two-fifths of the cost,
debt rises to 5.9 per cent of GDP above the early action scenario.

Potential for offsetting fiscal policy adjustments. Rather than increase total expenditure
to pay for the costs of decarbonisation, the Government could choose to allocate the
additional public investment from within its existing spending envelope. And rather
than allow existing taxes on motoring to fall to zero, the Government could maintain
the tax burden on motoring by levying other taxes such as a road-user charge. If net
zero investment were allocated from within the baseline, debt would be 8.4 per cent of
GDP lower in 2050-51 than if it were all additional (as it is in the early action
scenario). And if the tax burden on motoring were maintained (in contrast to the early
action scenario) it would be 24 per cent of GDP lower. Doing both would leave debt
32 per cent of GDP lower in 2050-51, which would actually be 12 per cent of GDP
lower than the hypothetical baseline, reflecting the gains from additional carbon tax
revenues.

13 Fiscal risks report

Executive summary

Chart 6: Climate scenarios: impact on public sector net debt in 2050-51

70
Denominator effect Net zero receipts losses
60
Carbon tax revenues Net zero public spending
50
Indirect effects Debt interest spending
40
Overall effect
Per cent of GDP

30

20

10

-10

-20

-30
Investment included and High productivity Early action Low productivity Late action
car taxes maintained variant scenario variant scenario
Source: OBR

Cost of public debt

24 Chapter 4 considers the fiscal risks associated with the Government’s elevated stock of debt.
This stock is both the product of past fiscal risks having crystallised and a source of future
fiscal risks that could. The risks that this poses to the fiscal outlook depend on the future path
of interest rates and the speed with which any changes in them are reflected in the
Government’s debt servicing costs. Over the past four decades, net interest payments by the
Government have fallen fourfold from 3.8 per cent of GDP in 1980-81 to 0.9 per cent in
2020-21, despite the debt-to-GDP ratio more than doubling from 40 to 100 per cent in that
time. This reflects the downward drift in global and UK short and long-term interest rates to
historically low levels, both in absolute terms and relative to the growth in GDP.

25 But higher post-pandemic government debt, combined with a shorter effective debt maturity
as a by-product of quantitative easing, leaves the UK’s public finances more exposed, and
more quickly, to increases in interest rates. The Government’s current fiscal plans, which
delivered a stable medium-term outlook for underlying public sector net debt as a per cent of
GDP in our latest forecast, were conditioned on rates remaining low, in line with market
expectations. But were they to return to historically more normal levels, it would become
significantly more expensive to service a given stock of debt. The Government acknowledged
this risk by making explicit reference to monitoring debt servicing costs in the fiscal targets
that guided the current Chancellor’s 2020 Budget.

26 As a starting point for our evaluation of the risk of future interest rate rises, we review the
various explanations that have been put forward for the decline in the cost of government
borrowing over the past thirty years. It is likely that demographic developments have played
some role, as well as slower productivity growth and increased preference for safe assets. But

Fiscal risks report 14

 Executive summary

there is considerable uncertainty about their respective contributions to the fall and their
permanence. While some of these factors are likely to remain in place in the future, justifying
our central forecast that borrowing costs will remain relatively low, others may reverse. Given
the uncertainty about the sources of the past decline, it is prudent to evaluate the risks to the
public finances were rates to rise. We therefore present several scenarios that illustrate the
consequences of assuming different future paths for borrowing costs, inflation, and GDP
growth for the public finances.

Higher global real interest rates

27 Our first two scenarios examine the fiscal impact of a globally driven rise in real interest
rates. In the first of these (‘higher R and G’), a gradual rise in real interest rates (‘R’) of 2.5
percentage points is associated with a similar pick-up in productivity growth (‘G’). In the
second scenario (‘higher R’), the higher interest rates occur not as a consequence of a pick-
up in productivity growth but as a consequence of a shift in investor preferences towards
riskier assets and away from government bonds.

28 In the first and more benign scenario, borrowing reaches 5.1 per cent of GDP in 2050-51
(versus 2.9 per cent in the baseline), pushed up by higher interest rates as net interest
payments rise to 3.3 per cent of GDP – a level last seen in 1985-86. But that is offset by
higher growth so that the debt-to-GDP ratio falls slightly below the baseline, although it
remains above pre-pandemic levels at the end of the scenario in 2050-51. In the second
and more challenging scenario, higher interest rates gradually feed through to the effective
interest rate paid on government debt, pushing borrowing up to almost 7 per cent of GDP in
2050-51. Without the offsetting gain from faster growth, debt hits 139 per cent of GDP by
2050-51, its highest level since 1954-55.

Higher inflation
29 Borrowing costs could rise not only because real interest rates rise but also because inflation
rises. The recent strong rebound in activity, expansionary macroeconomic policies (especially
in the US), and inflation outturns have prompted speculation of a reappearance of inflation.
We therefore consider two inflation scenarios. In the first, a burst of domestically generated
inflation of 5 per cent necessitates a temporary rise in Bank Rate to bring inflation back to
the 2 per cent target. In the second, consistent either with continued sanctioned inflation
overshoots or an increase in the target (as some commentators have advocated), inflation
runs persistently at 4 per cent, with a corresponding rise in short-term interest rates and
somewhat larger rise in long-term bond rates reflecting a higher inflation risk premium.

30 The fiscal implications of these scenarios demonstrate that inflation is, in fact, no longer a
very effective way to reduce the debt-to-GDP ratio, reflecting both the shortening of the
effective maturity of public debt as a by-product of quantitative easing and the relatively high
proportion of index-linked debt in the UK. The temporary burst of inflation has only a modest
impact on the debt-to-GDP ratio, which initially falls more quickly than the baseline mainly
due to primary spending being held constant in cash terms. By 2050-51, debt reaches 95
per cent of GDP, just 2 per cent of GDP below the baseline. With a persistent rise in inflation,

15 Fiscal risks report

Executive summary

there is a marginal improvement in the debt-to-GDP ratio in the first 13 years as inflation
erodes the real value of the nominal debt (though again moderated by the shortening of the
effective maturity of debt). But in the long run, the debt-to-GDP ratio actually rises to 107 per
cent of GDP by 2050-51 (10 per cent of GDP above the baseline) as a result of the extra
inflation risk premium being paid on the government’s borrowing.

A loss of investor confidence

31 Our final scenario explores the extreme case of a loss of investor confidence in the UK’s
creditworthiness that causes a flight from UK government bonds. This leads to a vicious circle
where rising debt raises borrowing costs, which in turn increase the rate at which debt rises.
In this scenario, an adverse shock, similar in magnitude to that experienced in the financial
crisis, and a loss of investor confidence lead to a sterling depreciation and a rise in the risk
premium on gilts. Higher inflation and the falling pound also force the Bank of England to
raise Bank Rate to 4 per cent. The higher borrowing costs mean that growth remains weak.
The escalating crisis also forces the Government to borrow at shorter maturities so that
higher market rates feed through into debt interest costs even more quickly.

32 Borrowing increases throughout the scenario due to a worsening primary balance as the
economy shrinks, as well as escalating interest costs. Borrowing reaches 15 per cent of GDP
in 2029-30, close to its peak last year. The adverse feedback loop between higher debt and
higher gilt rates leads to steadily rising interest costs, with the debt-to-GDP ratio increasing in
every year. By 2029-30, the average gilt rate hits 10 per cent – a rate last seen in 1991. We
end the scenario in 2029-30 when the government’s interest costs reach 9.5 per cent of
GDP, above any level seen in war or peacetime.

Chart 7: Cost of public debt scenarios: public sector net debt

180
Baseline
Higher R and G
160
Higher R
Temporary inflation shock
140
Persistently higher inflation
Loss of investor confidence
120
Per cent of GDP

100

80

60

40

20
2006-07 2011-12 2016-17 2021-22 2026-27 2031-32 2036-37 2041-42 2046-47
Source: ONS, OBR

Fiscal risks report 16

 Executive summary

Sensitivity to interest rate changes

33 These scenarios illustrate the greater sensitivity of public debt to future changes in its cost.
This is partly due to the trebling in the debt-to-GDP ratio since just prior to the financial crisis.
However, it is also the result of the shortening of the effective maturity of that debt as a by-
product of quantitative easing, which has replaced relatively long-dated gilts with (less
expensive) central bank reserves carrying an overnight rate of interest. The net result has
been a shortening of the median maturity of the consolidated liabilities of the public sector
from seven to two years since 2008 (the red line in the right panel of Chart 8). That contrasts
with the rise in the mean maturity of the total stock of gilts (in both public and private hands)
from 14 to 15 years over the same period. The proportion of debt on which interest rates
respond within a year has more than doubled over that time, which combined with the debt-
to-GDP ratio being almost three times higher, has made the first-year fiscal impact of a one
percentage point rise in interest rates six times greater than it was just before the financial
crisis, and almost twice what it was before the pandemic, just 18 months ago.

Chart 8: Sensitivity of interest payments to a rise in interest rates

1.2 26
Rise in interest payments from a one Maturities
24
percentage point rise in interest rates Mean of gilts
1.0 22
Total impact Mean of gilts, Treasury bills and NS&I
20
of which within one year Mean of public sector liabilities (net of APF)
0.8 18
Median of public sector liabilities (net of APF)
Per cent of GDP

16
14
Years

0.6
12
10
0.4
8
6
0.2 4
2
0.0 0
1998 2001 2004 2007 2010 2013 2016 2019 1998 2001 2004 2007 2010 2013 2016 2019
Note: Consolidated public sector liabilities are proxied here by the stock of Bank reserves, Treasury bills, NS&I products and gilts
net of those held in the APF. The total impact of a one percentage point rise in interest rates is on consolidated public sector
liabilities. The impact within one year is on liabilities with a maturity of less than 1 year (gilts net of APF holdings with a remaining
maturity of under one year, Bank reserves, Treasury bills and NS&I). The median shows the year in which half of the outstandi ng
public sector liabilities would be impacted by a change in interest rates.
Source: Bank of England, Heriot-Watt/Faculty and Institute of Actuaries Gilt Database, ONS, OBR

Other fiscal risks

34 While this report focuses on three large and looming threats to the public finances, the array
of other fiscal risks highlighted in previous FRRs has not evaporated. Indeed, as history warns
can happen, the pandemic has triggered the crystallisation of several of these risks,
aggravated many others, and even diminished a few. In Chapter 5 we set out how these
other risks have evolved since our previous FRR in 2019 and how our full risk assessment has
changed after factoring in both those changes, and the major risks discussed above.

17 Fiscal risks report

Executive summary

35 We have recast and consolidated some of the risks that were identified on our 2019 risk
register, so for this report we start from a total of 97 risks. Of these, we find that:

14 have crystallised including weaker productivity growth, lower net migration, and the
declining proportion of spending subject to firm DEL controls. Of these, 13 remain
active risks in future (including normal cyclical downturns, the deterioration in public
sector net worth, and cost overruns for major projects) and 1 has been removed (the
balance sheet risk relating to the classification of housing associations).

19 have increased, including those related to higher future health and social care
spending as a result of the pandemic, the longer-term sustainability of the fuel duty tax
base in light of the bringing forward of the ban on petrol-driven cars, and the
pandemic-driven increase in the non-payment of taxes due.

11 have decreased, including the tendency for fiscal policy to respond asymmetrically
to movements in our underlying forecasts following the tax rises announced in the
March Budget, the risks associated with persistent household financial deficits in light
of the savings accumulated by some during the pandemic, and the loss of revenue
from people moving to more lightly taxed forms of employment status.

29 remain unchanged, including our broad assessment around risks associated with
the financial sector which has so far weathered the coronavirus storm, clean-up costs
for nuclear plants, and those around stated policy aspirations.

3 have been resolved and removed from the register, including those around the
possibility of a ‘no deal’ Brexit and the rise in local authorities’ prudential borrowing
for commercial property purchases.

21 have been removed for other reasons including their being unquantifiable,
superseded by analysis presented in this report, or consolidated with other risks (taking
the total number of risks removed from the register to 25).

36 Finally, 15 risks have been added in this report, including nine arising from the coronavirus
pandemic, three associated with climate change, two relating to the cost of public debt and a
final one on the threat posed by a potential cyberattack. This takes the total number of risks
in our 2021 register to 87. Chart 9 depicts these changes as well as the number of risks that
have been affected to some extent by the pandemic.

Fiscal risks report 18

 Executive summary

Chart 9: OBR fiscal risk register: changes since our 2019 report
35

30

25
Number of risks

20

15

10

0
Crystallised Increased Unchanged Decreased Resolved Removed* Added
(ongoing)
Note: Darker shaded portions show the number of risks within each category that have been affected to some extent by the pandemic.
* The one risk that has crystallised and is no longer on the register is included in 'Removed'.
Source: OBR

37 Reflecting the correlated nature of fiscal risks, of the 97 risks from 2019, 38 have been
affected to some extent by the coronavirus pandemic. This includes around half of the
economy risks, two-thirds of the public spending risks, half of the risks relating to the
Government’s balance sheet and one-third of revenue risks. These include the pandemic-
related pressures on health spending and drop in net migration described above, as well as
the interaction between pandemic-driven fluctuations in earnings growth and the state
pensions triple lock that could cost £3 billion a year relative to our March forecast.

38 However, it is notable that one major and recurrent source of fiscal risks for the UK, that of a
financial crisis, has not crystallised despite the strains of the pandemic. This reflects both the
strengthening of capital requirements and other bank regulations since the financial crisis, as
well as the extensive and pre-emptive action taken by the Government and Bank of England
that protected household incomes, kept firms liquid, and maintained the supply of credit.

39 Alongside this report we have also published an updated and comprehensive risk register on
our website, listing all the fiscal risks discussed in this report, our assessment of their size and
likelihood, and, for those identified in our 2019 report, any changes since then. Figures 1
and 2 summarise the main risks to our medium-term fiscal forecasts and to long-run
sustainability respectively, categorised by size and likelihood.

19 Fiscal risks report

Executive summary

Figure 1: Sources of fiscal risk over the medium term

)

Change since 20
Igh ( ver 0)

evere
2 (per cent of G

recession
Added risk in 202
inancial crisis Typical recession

eak productivity growth

a or statistical utput gap
Impact on public sector net debt by 202

0)

reclassifications ( ) mismeasurement ( )
ower consumption and
labour income shares iscretionary fiscal loosening
edium (

igher interest rates ( ppt)

igher inflation ( ppt) ost pandemic pressures
ow migration on public services
igher cost of tax reliefs
Triple lock higher cost ( )
andemic scarring of economy

igher tax litigation costs

ow ( ess than )

ower consumption share

ower incomes for high earners Additional health
ower dutiable Additional adult social spending ( )
aster growth in self emplo y consumption ( ) care spending
ment and incorporations
o fuel duty R I increases
ower prices for expensive Government guaranteed loans
houses

ery low edium ow igh ery high

( ess than 0)( 0 0) ( 0 0)
( 0 0) ( ver 0)
robability of crystallisation (per cent)
ote 20 stress test no deal Brexit has been resolved and ow migration has crystallised and therefore have been removed.
denotes crystallised but remains active
ource BR

Figure 2: Sources of risk to fiscal sustainability

)

Change since 20
Igh ( ver 00)

ossible ma or fiscal
(per cent of G

risks that we have not

ealth spending other Added risk in 202
quantified Unmitigated climate change
a or wars cost pressures
Cyber attack

ower productivity growth

Impact on public sector net debt by 20 0

00)

ealth spending ageing

ow migration A typical recession
every decade
edium ( 0

igher growth corrected inancial crisis

Triple lock high cost
interest rate
Triple lock central
ate transition to net ero

uture pandemics
ow ( ess than 0)

Continued growth in Greater fuel efficiency

incorporations
Adult social care ageing
arly transition to net ero ower smoking

ery low ow edium igh ery high

( ess than 0) ( 0 0) ( 0 0) ( 0 0) ( ver 0)
robability of crystallisation (per cent)
ote e udge the igher growth corrected interest rate risk to have increased but has moved from high to medium impact due to improved scenario
estimations. denotes crystallised but remains active
ource BR

Fiscal risks report 20

 Executive summary

Lessons for understanding fiscal risks

40 Our foregoing analysis of the experience of the pandemic, the threat posed by climate
change, and the risks to the cost of public debt, points to 10 lessons for understanding and
responding to potentially catastrophic fiscal risks:

1 Catastrophic risks are real and may have become more frequent. Just two decades
into this century, advanced economies have now experienced two ‘once in a century’
economic shocks. And increasing economic and financial interconnectedness may
make future shocks both more frequent and more severe. Putting greater emphasis on
the analysis of risks and the uncertainty surrounding our central projections will help
ensure that policymakers can incorporate these risks into their decision making and
the public understand the trade-offs being made.

2 Economic shocks affect both supply and demand. While conventional cyclical shocks
affect mainly demand, recent shocks – the financial crisis, Brexit and the pandemic –
have each hit both supply and demand. The principal focus of the Government’s
coronavirus rescue package was the preservation of supply-side capacity while
demand was deliberately suppressed. Tackling climate change requires action to
address not only the (excessive) demand for carbon but also (inadequate) supply of net
zero technologies. And understanding risks to the cost of government debt requires
investigation of the drivers of supply and demand for gilts. We need to improve our
understanding and modelling of the supply-side impacts of shocks and policies.

3 Global interconnectedness can be both an asset and a liability. The UK’s openness
exposes it to risks emanating from abroad, but it also attracts the international talent
and investment that has made it a world leader in genomic sequencing and vaccine
research and development. Digital connectivity enabled our economy to continue to
operate through the pandemic, but also renders it vulnerable to cyberattacks. And
rising overseas demand for UK government debt has helped to keep gilt yields low, but
also exposes the public finances to sudden changes in international investor sentiment.

4 While it is difficult to predict when catastrophic risks will materialise, it is possible to

anticipate their effects. While a global pandemic topped government risk registers for
over a decade before coronavirus arrived, it attracted little attention from the
economics community. But previous epidemics and pre-pandemic modelling by official
bodies provided a clear indication of how economies might be affected. Similarly, the
tail risks associated with extreme climate change or spiralling debt servicing costs
scenarios can be quantified, even if we do not know precisely when they might occur.

5 When investing in risk prevention, governments have a tendency to ‘fight the last war’.
The regulatory response to the 2008 financial crisis helped prevent the pandemic from
triggering another financial crisis. But post-crisis fiscal consolidation also cut advanced
country expenditure on preventative health programmes. Dealing with post-pandemic
economic and fiscal pressures may hamper governments’ efforts to invest the relatively
modest sums need to avoid the much greater cost of unmitigated climate change.

21 Fiscal risks report

Executive summary

More regular and comprehensive horizon-scanning could help to identify where the
next crises could emerge and how they can be prevented or mitigated.

6 There are advantages in preventing or halting a process that involves rapidly

escalating costs early. While economic theory and practice emphasises the option
value of delaying decisions, this can be suboptimal in the face of rapidly rising costs.
Pandemics, climate change, and public debt dynamics are all subject to amplifying
feedback mechanisms and tipping points that can result in spiralling and irreversible
costs that put a premium on acting early. Countries that acted quickly to contain the
virus have so far experienced fewer deaths, shallower recessions, and faster economic
recoveries. In making the transition to net zero, delaying decisive action to tackle
carbon emissions by ten years could double the overall cost.

7 People appear willing to make sacrifices for a clearly defined public good. Levels of
compliance with public health restrictions and vaccine take-up in the UK have been
surprisingly high. In total, the UK experienced a 10 per cent loss of output and
committed 12 per cent of GDP in public funds in order to combat the pandemic in
2020. The annual economic and fiscal costs of tackling other potential catastrophic
risks, like climate change, are likely to be just a fraction of this.

8 Economies can sometimes adapt remarkably quickly to structural changes. While the
initial output loss from lockdowns was greater than many predicted, most were also
surprised by the speed at which the economy adapted and rebounded as restrictions
were lifted. Prior investments in digital infrastructure and services were critical to
enabling this transition, as was fiscal support to households and firms. The path to net
zero will also require more gradual adaptation to new technologies and changes in
behaviour. But fiscal policy could also play a role in facilitating the transition.

9 Fiscal policy can and needs to be more nimble than in the past. Across advanced
economies the fiscal policy response to the pandemic was unprecedented in its speed,
scale, and novelty (partly reflecting constraints on monetary policy). This added over
20 per cent of GDP to debt, but also prevented the much greater economic costs
associated with not intervening. Similar fiscal policy nimbleness and creativity may be
required to support an economy-wide transition to net zero. And flexibility in both
deploying, and withdrawing, fiscal support is likely to be critical if governments are to
respond to future shocks without jeopardising debt sustainability in the long run.

10 In the absence of perfect foresight, fiscal space may be the single most valuable risk
management tool. Throughout its history, the UK has relied on its ability to borrow
large sums quickly in order to respond to major economic and political threats. It was
able to do so courtesy of its relatively low levels of public indebtedness, deep and
liquid domestic capital markets, and by maintaining the confidence of international
investors in its long-run creditworthiness. In the face of an array of major economic
and fiscal risks, policymakers must trade off making significant investments in the
prevention of specific potential threats with preserving sufficient fiscal space to respond
to those risks that it did not anticipate or could not prevent.

Fiscal risks report 22

1 Introduction

Background
1.1 The UK has been a leader in the analysis of fiscal risks in recent years.1 The legislation
establishing the OBR has always required us to set out the main risks that we consider to be
relevant in any report that we produce.2 From our establishment in 2010, our biannual
Economic and fiscal outlooks (EFO) have regularly featured fan charts, sensitivity analysis,
and alternative scenarios to illustrate the risks around our central forecasts. The
September 2015 review of the OBR expanded our remit in this area by recommending that
we produce a new report on fiscal risks, extending existing analysis and meeting the
3

Parliament reflected this in the October 2015 edition of the Charter for Budget
Responsibility. And the November 2016 Fiscal Transparency Evaluation of the UK
subsequently recommended that the report should represent a comprehensive and
quantified fiscal risk statement that includes all major risks to the fiscal position 4

1.2 The Charter tasks us with producing a biennial re the main risks to the public
finances, including macroeconomic risks and specific fiscal risks We published our first
Fiscal risks report (FRR) in July 2017. Several countries produce regular fiscal risk
assessments, but most are undertaken by finance ministries or cabinet offices. The UK is
unusual in outsourcing it to an independent fiscal institution, thereby boosting objectivity
and transparency in the analysis of fiscal risks. And the UK is also unique in setting a legal
requirement for the Treasury to respond formally to our FRR within a year of its publication,
thereby encouraging accountability for .

1.3 This chapter describes our approach to analysing and reporting on fiscal risks. It starts by
defining fiscal risks and distinguishing between those risks that governments are exposed to
in normal times and the large, and potentially catastrophic, shocks that governments face
from time to time. It then considers whether the latter type of risk may be becoming more
relevant for advanced economies in the twenty-first century, and how this motivates the
content of our third FRR.

1
This has been recognised by both the IMF and the OECD. See, for example, The UK Fiscal Risk Report raises the bar on the assessment
and quantification of fiscal risks to a new level in Stressing the public finances the UK raises the bar, IMF Public Financial Management
blog, July 2017, and, OECD, Independent Fiscal
Institutions Review of the OBR, September 2020.
2
Section 4, subsection 6(b) of the 2011 Budget Responsibility and National Audit Act.
3
HM Treasury review of the Office for Budget Responsibility, Led by Sir Dave Ramsden, Chief Economic Adviser to HM Treasury, HM
Treasury, September 2015.
4
United Kingdom: Fiscal Transparency Evaluation, IMF, November 2016.

23 Fiscal risks report

Introduction

Our approach to analysing fiscal risks

What is a fiscal risk?
1.4 The Interna the possibility of deviations of
fiscal outcomes from what was expected at the time of the Budget or other forecast 5 On
this basis, we would define a fiscal risk as any potential deviation from the 5-year-ahead
central forecasts for public sector spending, receipts, borrowing and debt contained in our
EFOs, and from the corresponding 50-year-ahead projections in our Fiscal sustainability
reports (FSR). We are required by Parliament to base these forecasts and projections on
currently stated Government policy, although in most cases current policy is much less
clearly defined over the long term than over the medium term. Where appropriate, we
consider policy risks areas where government statements or past behaviour point to likely
future policy changes (as we do throughout this report).

1.5 On this definition, however, what constitutes a fiscal risk depends crucially on which
developments in the public finances are incorporated into our central projection and which
are regarded as potential deviations. Given the sensitivity of long-term projections to these
sorts of judgements, we focus on risks around our central forecast over the medium term,
but on risks to fiscal sustainability (rather than around our latest central projection) over the
longer term. This ensures that we do not end up ignoring some of the most important long-
term risks notably pressures on health spending simply because we already assume they
crystallise gradually over time.

1.6 Our focus on risks to sustainability also implies an asymmetry of approach we are more
interest
countries suggests that shocks to the public finances (especially big ones) are more likely to
be adverse than beneficial as the cost of the coronavirus pandemic has illustrated so
dramatically and that governments are usually quicker to spend unexpected windfalls from
good news than they are to anticipate and provision for unexpected costs from bad news.

1.7 The definition of fiscal risk we use in this report focuses on surprises relative to forecasts and
pressures on fiscal sustainability. But it is important to remember that the purpose of much
of government activity is to pool risks that society has decided (via the political process)
would be better carried by the state than borne by individuals (either directly or through
private insurance markets). For example:

During normal times, the state provides a degree

For example, the NHS takes on the health costs people would otherwise face when
they fall ill; state pensions put a floor under
reduces the risks associated with periods of unemployment or on low pay.

During catastrophes and other crises, states often take on a much broader range of
costs as large risks crystallise. For example, during the current pandemic the

5
IMF Fiscal Affairs Department, Fiscal risks sources, disclosure and management, 2009.

Fiscal risks report 24

 Introduction

Government has at some point paid the salaries of over 8.7 million furloughed staff;
guaranteed loans to 1.6 million businesses; and provided grants to support the
incomes of 2.6 million self-employed people.

Our first two Fiscal risks reports

1.8 Our first two FRRs considered both sorts of risk, devoting much of their analysis to the risks
facing government in normal times while noting that periodic shocks were inevitable:

FRR 2017 attempted a comprehensive survey of the universe of risks to the public
finances, ranging from whole economy risks emanating from macroeconomic shocks
and financial crises down to the long tail of generally smaller risks to individual
components of the public finances (from the potential costs of reforming adult social
care to the potential loss of fuel duty receipts as cars become more fuel efficient). It
also included a fiscal stress test based on a severe recession scenario used by the Bank
of England to assess the financial resilience of the UK banking sector. We identified 57

Managing fiscal risks publication, the Government detailed its approach to these issues
and the steps it had taken in several areas to enhance its risk management.

FRR 2019
response, which allowed us to assess the degree to which risks had intensified or
abated. We also looked more deeply into several key risks that had been covered in
less depth, or not at all, in our first report, including: fiscal policy risks, -
corrected interest and climate change. The report included another fiscal stress
test, this time based on an IMF no-deal Brexit scenario. Our report again raised a set
of issues for the Government to consider in its response, but this was overtaken by the

FRR in July 2020 understandably constituted a brief written statement to Parliament.6

This Fiscal risks report

1.9 Unlike previous editions, this third FRR focuses not on the risks that the government faces in
normal times but the exceptional, systemic shocks that can potentially lead to economic and
fiscal crises. The reasons for doing so are self-evident. The coronavirus pandemic has
provided a stark reminder that catastroph , while inherently more difficult to
anticipate and analyse, are real. These include financial crises, severe recessions, extreme
weather events, destructive cyberattacks, pandemics, and major armed conflicts. While most
advanced economies have been largely spared such catastrophic risks in the latter half of
the twentieth century (and certainly by comparison with the first half),7 there are several

6
Chancellor of the Exchequer, OBR 2020 Fiscal Sustainability Report and response to the OBR 2019 Fiscal Risks Report, 15 July 2020.
7
This is not to say that advanced economies did not experience any shocks in the second half of the twentieth century. The Korean War in
the late 1940s and early 50s; the Vietnam War in the 1960s and 70s; the oil shocks of the 1970s; the bursting of the Japanese bubble in
the late 1980s; Black Wednesday in the UK, and the Scandinavian and Asian financial crises of the 1990s, all had significant adverse
economic and fiscal impacts on the economies most directly affected. But none of these matched either the 2008 financial crisis or 2020
coronavirus pandemic in the depth of the fall in global output or number of countries adversely affected.

25 Fiscal risks report

Introduction

reasons to believe that they are increasingly susceptible to large and disruptive economic
and fiscal shocks in the twenty-first. The next section explores why this might be so.

1.10 This edition of the FRR does not include a new scenario-based fiscal stress test, but the
discussion of the fiscal impact of the pandemic in Chapter 2, and the more severe climate
change and debt scenarios in Chapters 3 and 4, serve the purpose that stress tests have in
our previous FRRs. And the update on other fiscal risks in Chapter 5 underscores one of the
central insights that those stress tests provided that fiscal risks are highly correlated and
governments can face a cascade of crystallising risks when hit by large shocks.

Is the world becoming a riskier place?

1.11 So far this century, the UK has been hit by in the
form of the 2008 financial crisis and 2020 coronavirus pandemic. While two observations
do not constitute a trend, these events raise important questions for fiscal forecasters and
policymakers about the nature of risk in the twenty-first century and how to respond to them
specifically: Are catastrophic risks becoming more likely? Are catastrophic risks becoming
more severe? Are countries becoming more exposed to catastrophes happening elsewhere?

Incidence of catastrophic risks

1.12 The first two decades of this century have shown that major shocks to advanced economies
can emanate from a variety of sources. The incidence of some of these shocks, like the
earthquakes in Japan and New Zealand, depend on natural forces outside human control.
The incidence of other major risks, such as cyberattacks, are entirely the product of human
action. For many risks, such as pandemics and climate change, the incidence of shocks is a
product of the interaction of people with their environment. Compared to the previous
century, at least one important source of potentially catastrophic risk, that of armed conflict
between states (especially nuclear powers and their close allies), appears to have
diminished (Chart 1.1).8 That said, deaths from civil conflicts within states (including with
foreign state intervention) have remained significant throughout the first two decades of this
century.9 And such conflicts can put pressure on advanced economies by generating large
refugee flows and providing havens for international terrorist groups.10

8
Compared with the first half of the twentieth century, the decline in deaths due to inter-state conflicts is of course much greater. Indeed,
more battle-related deaths were recorded during the two World Wars than were recorded in the entire period since 1946.
9
Deaths have been largely concentrated in the Middle East in the past decade.
10
HM Government, National Risk Register, 2020 Edition, 2020.

Fiscal risks report 26

 Introduction

Chart 1.1: Battle-related deaths in state-based conflicts

25
Colonial or imperial conflicts

Civil conflicts with foreign state intervention

20
Civil conflicts

Conflicts between states

Deaths per 100,000

15

10

0
1946 1951 1956 1961 1966 1971 1976 1981 1986 1991 1996 2001 2006 2011 2016
Source: International Peace and Research Institute, Oslo; Uppsala Conflict Data Programme

1.13 At the same time, recent surveys of systemic risks to the global economy point to growing
threats from other sources. Even before the coronavirus pandemic, the global risks
produced by Cambridge University pointed to a steady increase in
from a range of sources, putting the total at $577 billion (1.6
per cent of world GDP) in 2019.11 Their most recent report identified financial crises,
interstate conflicts, climate change, human pandemics, and cyberattacks among the top
threats (largely as set out in the 2020 edition of
its National Risk Register12). This partly reflects a rise in the frequency of risks materialising,
especially in the case of severe weather events and human infectious disease outbreaks,
where the numbers of reported incidents have doubled and trebled respectively since the
1990s (Chart 1.2). This may partly be a function of increased surveillance and reporting of
incidents. However, it is also likely to be driven by the growing numbers of people living
closer to the sources of risks such as flood plains and isolated ecosystems.13

11
University of Cambridge Judge Business School, Cambridge Global Risk Index 2019, 2019. GDP-at-risk sums estimates across 279
major cities around the world (covering 41 per cent of world GDP) that are based on their economic output, their exposure to particular
threats related to their geography and type of economy, offset by their estimated resilience in recovering from shocks.
12
HM Government, National Risk Register, 2020 Edition, 2020.
13
Gavi, How has our urban world made pandemics more likely?, 2020.

27 Fiscal risks report

Introduction

Chart 1.2: Numbers of international disasters and infectious disease outbreaks

500 800
Climatological Outbreaks originated in humans
450 700
Geophysical
400 Outbreaks originated in animals
Meteorological 600

Number of outbreaks
350
Number of events

Hydrological 500
300
250 400
200 300
150
200
100
50 100

0 0
1980 1985 1990 1995 2000 2005 2010 2015 1980 1985 1990 1995 2000 2005 2010

Source: The Emergency Events Database Source: KF Smith et al. (2014)

Severity of catastrophic risks

1.14 The rise in the amount of global GDP at risk from catastrophes also reflects an increase in
the severity of some of those incidents when they materialise. For example, hurricanes have
not only become more frequent in the North Atlantic over the past fifty years, with 2020
seeing the highest ever number of named storms at 29, but also more destructive as global
temperatures rise.14 In the case of risks emanating from the economy, studies of business
cycles among major economies suggest that, while periods of economic expansion among
G7 economies have become smoother and longer, economic contractions have become
larger and more severe than preceding expansions. This implies that risks arising from
macroeconomic volatility have become increasingly skewed toward the downside with
longer periods of steady economic expansion punctuated by deeper recessions.15 This partly
reflects greater financial integration both within and across countries, which allows for
longer credit cycles but also gives rise - .

Transmission of catastrophic risks

1.15 Related to the above, a final reason why the world may be becoming riskier is the increase
in global interconnectedness. The past forty years has witnessed a seven-fold increase in air
passenger numbers,16 twenty-five-fold increase in international capital flows,17 and a six-fold
increase in the volume of international trade.18 So even in areas where the incidence and
severity of risks has stayed the same or declined, the potential for those risks to be
transmitted between countries has risen, increasing the risk exposure for any given
country.19 This is especially relevant for countries like the UK which, according to one index
of cross-border economic linkages compiled by DHL, is the ninth most globally connected
country in the world (Chart 1.3).20

14
Li, L., and Chakraborty, P., Slower decay of landfalling hurricanes in a warming world, Nature, Vol. 587, 2020.
15
Jensen, H., et al, Leverage and Deepening Business-Cycle Skewness, American Economic Journal: Macroeconomics, Vol. 12, No.1, 2020.
16
World Bank, Air Transport, passengers carried, 2021.
17
IMF, Balance of Payments and International Investment Position Statistics, June 2021.
18
World Bank, Exports of goods and services (constant 2010 US$), 2021.
19
Ferguson, N. Doom: The Politics of catastrophe, 2021.
20
DHL, Global Connectedness Index 2018: The State of Globalisation in a Fragile World, 2018.

Fiscal risks report 28

 Introduction

Chart 1.3: Global connectedness

Are governments more exposed to catastrophic risks?

1.16 As advanced economies exposure to potentially catastrophic risks increases, so do the

associated risks to the public finances. This is not only because of the disruptive effects of
the associated economic shocks on government revenues and non-discretionary spending.
Governments are also more directly exposed because catastrophic risks are, by their nature,
difficult or impossible to price or insure against. This means that the private sector cannot
manage them without active government intervention. Government is therefore in effect
obliged to step in to act as an . This was the topic of a timely Treasury
report published in March 2020, which set out a series of proposals for improving the
management of government guarantees and other contingent liabilities.21

1.17 However, the experience of the past t

of its insurer of last resort function goes beyond just the issuance of guarantees and other
explicit forms of insurance. During the financial crisis, advanced economy governments
stepped in to acquire or underwrite the assets of their largest banks in order to prevent an
even greater credit crunch. During the coronavirus pandemic, governments stepped in to
provide grants to businesses and help pay the wages of individuals hit hardest by the
pandemic. These and other interventions in the economy dramatically increased
government cash outlays at the same time as output and revenues were squeezed.

21
HM Treasury, Government as insurer of last resort: managing contingent liabilities in the public sector, 2020.

29 Fiscal risks report

Introduction

1.18 The fiscal risk associated with governments costs in the event of
major shocks may be increasing over time. This partly reflects the more disruptive nature of
recessions in an era of deeper financial integration and greater financial leverage, as
discussed above. Significant government interventions in asset and credit markets are
therefore required to prevent widespread bankruptcies or to avoid disorderly workouts of
corporate and household debts. More recently, it reflects the limitations on the ability of
monetary policy to support the economy imposed by the effective lower bound on interest
rates. It is also due to the idiosyncratic nature of the coronavirus shock, which required the
kind of targeted intervention in the most affected sectors that only fiscal policy can provide.

Challenges for fiscal policymakers

1.19 The fact that fiscal risks may be becoming more frequent, severe, and contagious,
combined with increasing expectations that government will bear the immediate costs of
shocks, poses important challenges for those managing the government finances. Both the
financial crisis and the pandemic have required governments around the world to exercise
their insurer of last resort function with dramatic and lasting consequences for government
borrowing and debt. Policymakers face a difficult trade-off in deciding how much to spend
today to reduce the odds of these potentially catastrophic risks from materialising versus
e their effects when they do. That
judgement is further complicated by the challenge of assessing quite how much fiscal space
is available to policymakers (Box 1.1).

Box 1.1: Assessing fi

Major fiscal shocks are by their nature varied, hard to predict, and difficult or very expensive to
mitigate entirely. Because no two shocks are the same, responding to them when they crystallise
typically requires the rapid and innovative deployment of government resources on a large scale

the room for undertaking discretionary fiscal policy relative to existing

plans without endangering market access and debt sustainability a is central to managing fiscal
risks. But how can this be translated into a practical guide for policy makers?
In its 2018 Managing fiscal risks report, the Treasury employed an OECD frameworkb to review

generally do not allow for more granular factors that

manoeuvre such as average debt maturities, share of inflation-linked debt, currency of debt
issuance, whether bondholders are mostly domestic or foreign, holdings of off-setting liquid
assets, extent of non-debt or contingent liabilities, and capacity to adjust fiscal policy to
accommodate rising interest costs. Indeed, the IMF has adopted a multi-faceted approach in
which many factors are considered, estimates are allowed to vary across country and time, but
which yields only a qualitative assessment.c This reflects the importance of taking a broad view of
the factors determining fiscal space, the challenges in quantifying some of them, and difficult

Fiscal risks report 30

 Introduction

judgements required in assigning relative weights to those factors for the purpose of coming up
with an overall quantified assessment of fiscal space.
Chart A: Estimates of government debt limits, thresholds, and targets
250
Prudent target
Debt threshold
200
Debt limit
Per cent of GDP

Latest outturn
150

100

50

0
OECD advan ced

OECD advan ced

EU stability and

Moody's UK
estimate (lower

estimate (higher
(1997-2008)

OECD UK
Ghosh et al.
UK Government
Analysis Framework

(2013) UK

estimate
(upper bound)
(lower bound)

(2015)
growth pac t

estimate
UK target

risk) (2015)

estimate
debt outturn
risk) (2015)
OECD UK

(2011)
OECD UK

(2020-21)
econ omies

econ omies
Sustainability
IMF Debt

Source: Ghosh et al. (2013), HMT, IMF, OECD, ONS

However, for many determinants of fiscal space the sign of the impact is clear even if the scale is
uncertain. All else equal, fiscal space increases with: lower levels of debt (so the past decade has
currency (a strength for the UK); a
longer maturity of debt (also a strength, although one complicated by the effects of quantitative
easing, as discussed in Chapter 4); holdings of high quality liquid assets; lower non-debt
liabilities such as unfunded pension obligations and contingent liabilities such as guarantees; a
capacity to rapidly adjust fiscal policy in response to shocks; and a track record of meeting debt
obligations (another UK strength). All these factors can vary over time and across countries.
The availability of fiscal space will also depend on the nature of the shock to which policymakers
are responding. For a common shock, such as the pandemic, countries with an established
reputation for meeting their obligations and whose bonds are traded in deep and liquid markets

affected. For governments whose debt is considered a safe haven, fiscal space can be highly
elastic as risk appetite shrinks, the demand for relatively safe assets increases and so too does
the availability of willing lenders to those safe havens, which increases their fiscal space to
borrow and respond to the shock.

But continued safe-haven status cannot be guaranteed and the cost of losing it can be
significant. In the face of an idiosyncratic shock, governments particularly those reliant on
foreign investors can see funds drain away into safer assets in unaffected countries, resulting in

31 Fiscal risks report

Introduction

higher borrowing costs and a reduction in fiscal space at precisely the moment the government
most needs it.
concluded
some fiscal space credible medium-term fiscal framework and a
credible fiscal consolidation plan would be needed.
a
-Press Release; Staff Report; Staff Supplement; and Statement by the Executive

b
Falilou Fall, Debra Bloch, Jean- OECD
Economic Policy Paper Number 15.
c
ce IMF Staff Paper
IMF Policy Paper.

Challenges for fiscal forecasters

1.20 ks also raises
challenges for fiscal forecasters such as the OBR. This challenge can be illustrated by
looking at the path of public sector net debt in the UK since the turn of the century. Debt
stood at a historically low 27 per cent of GDP in 2000-01 but is expected to reach 107.4
per cent of GDP in 2021-22 in our latest forecast. Of that 80 per cent of GDP increase,
two-thirds (just over 50 per cent of GDP) occurred in just four years at the height of the
financial crisis in 2008-09 and 2009-10 and at the height of the pandemic in 2020-21 and
2021-22 (Chart 1.4). Unsurprisingly, neither of these shocks were anticipated in either the
pre-OBR forecast prepared by the Treasury in March 2008 or our pre-pandemic forecast in
March 2020.

Chart 1.4: Public sector net debt

140
March 2008 forecast 2008 2020
financial coronavirus
March 2020 forecast crisis pandemic
120
March 2021 forecast

100
Per cent of GDP

80

60

40

20

0
2000-01 2002-03 2004-05 2006-07 2008-09 2010-11 2012-13 2014-15 2016-17 2018-19 2020-21 2022-23 2024-25
Source: ONS, OBR

Fiscal risks report 32

 Introduction

1.21 The tendency to significantly under-forecast debt is by no means a UK-specific

phenomenon. A review of IMF and Economist Intelligent Unit projections for 147 countries
over the past two decades found that these independent forecasters on average
underpredicted the level of debt five years ahead by about 10 per cent of GDP. Consistent
with
associated with surprise recessions in the forecast horizon.22 Occasional large or
catastrophic risks present a particular problem for forecasters, as it would make no sense to
assume they are always just about to happen, as most of the time they do not. But they do
crystallise sometimes and moreover on an unpredictable basis. That means that the analysis
and communication of risks, rather than just central forecasts, is key to providing a
complete view of economic and fiscal prospects.

1.22 As a further contribution in this area, we therefore hope to enhance the presentation of
uncertainty in future EFOs through the use of stochastic simulations. These involve
producing multiple scenarios that are driven by randomly selected shocks of the sort that
have been experienced in the past, so highlighting the distribution of risks around our
central forecast. This approach is employed by a number of organisations, including the
public debt sustainability. We will
set out our own intended approach in a forthcoming working paper.

Structure of this report

1.23 Against this background, our third FRR shifts the focus of our analysis onto three larger, and
potentially catastrophic, sources of fiscal risk: the coronavirus pandemic, climate change,
and the cost of government debt. These three risks emanate from different sources and have
their own unique drivers, but they also share some commonalities:

there is a significant degree of uncertainty concerning both the timing and the scale of
their associated costs;

they are characterised by non- in which costs can

escalate dramatically from the point of crystallisation, with potentially catastrophic
consequences for economies and public finances; and

they may be global in nature with high potential for rapid contagion of risk across
countries.

1.24 Governments seeking to manage these threats must therefore weigh the known costs of
early action to mitigate these risks against the uncertain costs of dealing with them if they
crystallise. And they need to weigh the limited but more deliverable benefits of acting
unilaterally against the much greater but more elusive gains from acting globally.

22
CEPR Discussion Paper No. DP16108, Worse than You Think: Public Debt Forecast Errors in Advances and Developing Economies, 2021.

33 Fiscal risks report

Introduction

The coronavirus pandemic

1.25 Neither we, nor most other fiscal analysts, saw the enormous economic and fiscal
consequences of a global pandemic coming. Our 2017 FRR note
2015 National Register of Civil Emergencies had identified an influenza pandemic as the
most significant civil emergency risk and the risk that such an event might pose to health
spending, but we did not attempt to quantify the broader fiscal risk that it might pose. The
US Congressional Budget Office did look at such risks back in 2005 (in a paper produced
at the behest of a Senator with a background in medicine23). It focused on an outbreak of
avian influenza and drew on the 1918 flu pandemic to calibrate scenarios that have proved
remarkably accurate in anticipating the costs of the coronavirus pandemic.

1.26 The economic and fiscal shock associated with the pandemic provides a classic example of
crystallising one whose impact is so large and whose likelihood in any given
year is so small that it sits in the very tail of the distribution of possible bad outcomes. The
shock to the y was the largest in over three centuries, since the Great Frost of
1709, and the resulting fiscal deficit was the largest the UK has witnessed in peacetime.

1.27 As a case study in the crystallisation of a catastrophic risk, Chapter 2 therefore explores: the
impact of the pandemic on the UK economy and public finances in historical and
international context; the economic and fiscal support extended by governments in response
to the pandemic; the legacy risks that the pandemic poses for the public finances over the
medium term; the potential longer-term implications of the pandemic for the economy and
public finances; and the lessons that the pandemic carries for understanding other
potentially catastrophic risks.

Climate change
1.28 Looking ahead, the catastrophic threat posed by unmitigated global warming and climate
change is clear. Governments around the world have recognised this and signed up to the
2015 Paris Agreement that seeks to limit global warming to well below 2 (preferably to 1.5)
degrees Celsius above pre-industrial levels. In the UK, the Government has since legislated
to achieve net zero greenhouse gas emissions by 2050.

1.29 The UK alone cannot affect the path of global warming to a material extent we accounted
for just 1 per cent of global emissions in 2019. So the catastrophic fiscal risks associated
with a global failure to meet the Paris targets are beyond the UK
tters the US, China and the EU all setting objectives to get to
net zero emissions, we can focus more narrowly on the fiscal risks posed by different paths
to net zero in the UK. Reflecting on the similarities and differences between the response to
both climate change and the pandemic prompted one study to conclude that The climate
emergency is like the COVID-19 emergency, just in slow motion and much graver. 24

23
R. Arnold, J. De Sa, T. Gronniger, A. Percy and J. Somers, A Potential Influenza Pandemic: Possible Macroeconomic Effects and Policy
Issues, US Congressional Budget Office, December 2005 (revised July 2006).
24
Will COVID-19 fiscal recovery packages accelerate or retard progress on climate change?
Nicholas Stern, Joseph Stiglitz, Dimitri Zenghelis, Oxford Review of Economic Policy, May 2020.

Fiscal risks report 34

 Introduction

1.30 In Chapter 3 we therefore consider the potential economic and fiscal consequences of:
unmitigated global warming; the array of policy levers available to support the
decarbonising of t
target for bringing greenhouse gas emissions in the UK down to net zero by 2050.

Cost of public debt

1.31 The pandemic has caused public debt to jump in the UK and around the world. On
average, the IMF expects advanced economies to see gross debt to GDP ratios at the end of
the 2021 that are 18.7 percentage points higher than on the eve of the pandemic in
2019.25 We forecast public sector net debt to rise from 80.4 per cent of GDP at the end of
2018-19 to 107.4 per cent of GDP at the end of 2021-22. Despite that increase in the stock
of debt, debt interest spending is expected to fall from £37.5 billion in 2018-19 to £24.8
billion in 2021-22 (from 1.7 to 1.1 per cent of GDP) thanks to falls in interest rates and the
near-doubling of quantitative easing.

1.32 Financial market participants currently expect interest rates to remain very low for the
foreseeable future, so at face value the stock of debt is cheaper to service despite the large
increase due to the pandemic. But higher debt also increases the sensitivity of the public
finances to movements in interest rates, so fiscal risks associated with the public debt have
risen. These risks are not so much that of a default on government debt the UK
government borrows in its own currency, has an independent central bank, and has an
enviable track record of honouring its debts. Rather, as Kenneth Rogoff has put it, the
problem of carrying very high public debt is not sustainability, but loss of flexibility in
responding to unforeseen shocks. 26 in a century
little over a decade, it is clear that medium- and longer-term fiscal prospects are contingent
on the ability of the Government to respond flexibly as and when shocks hit.

1.33 In Chapter 4 we therefore: look at the historical drivers of debt levels and interest rates;
assess the sensitivity of the public finances to alternative scenarios for the future path of
interest rates, inflation, and growth; and explore the fiscal consequences in the more
extreme scenario of a loss of investor confidence in UK government debt.

Update on other fiscal risks

1.34 Finally, the crystallisation of one major risk does not mean others have evaporated. Indeed
economic shocks can trigger the crystallisation of other fiscal risks. In Chapter 5 we look
back at the 98 risks that were identified in our 2019 FRR. We describe where and why our
assessment has changed, including the role the pandemic has played in those changes. We
then add those risks identified in this FRR to the 2021 edition of our fiscal risks register.

25
IMF, World Economic Outlook, April 2021.
26
Is Higher Debt an (Almost) Free Lunch?, Kenneth Rogoff, Harvard University, paper for the European Fiscal Board, February 2021.

35 Fiscal risks report

Fiscal risks report 36
2 Coronavirus pandemic

Introduction
2.1 In addition to being the most acute public health crisis the world has faced in over a
century, the coronavirus pandemic resulted in the largest economic shock that the UK has
experienced in three centuries.1 fforts to mitigate its impact on
businesses, households, and public services also prompted the most dramatic expansion in
the size and scope of government activity and the largest budget deficit in peacetime. The
global pandemic is still far from over, with worldwide coronavirus cases still averaging over
300,000 a day and economies continuing to face potential risks from new and more
vaccine-resistant variants.2 And even after the immediate public health risks have abated,
the pandemic may leave behind a legacy of medium-term pressures on public services and
long-term scars on the economy. However, eighteen months on from the start of the
pandemic, one can begin to draw some preliminary lessons from UK and international
experience of the pandemic for how to understand and manage other potentially
catastrophic fiscal risks such as those explored in other chapters of this report.

2.2 This chapter therefore explores:

the economic and fiscal impact of the pandemic, setting it in both historical and
international context;

the role played by fiscal policy in mitigating the immediate impact of the pandemic on
the economy and public finances;

the direct medium-term fiscal pressures left behind by the pandemic and the
government s policy response;

the indirect longer-term fiscal risks that could arise from the impact of the pandemic
on the supply side of the economy; and

some initial economic and fiscal lessons from the pandemic for how economic
forecasters and policymakers should approach other potentially catastrophic risks.

1
The coronavirus pandemic has resulted in just under 4 million deaths in 18 months and has been the most disruptive to the global
economy over the past century, but it is not the most deadly over this period. HIV/AIDS has killed 30-35 million people and smallpox
eradicated in 1980 is estimated to have caused around 300-500 million deaths over a century. The 1918 flu pandemic a little over a
century ago is estimated to have killed around 50 million people worldwide.
2
Daily new confirmed COVID-19 cases, World, Our world in data, 28 June 2021.

37 Fiscal risks report

Coronavirus pandemic

Economic and fiscal impact of the pandemic

2.3 The coronavirus pandemic was different from any other shock that the UK economy and
public finances have faced in peacetime. Unlike previous major economic downturns, the
source of the disturbance lay outside the economy and its management, and its resolution in
the first instance required public health interventions rather than any change in economic
policy. Indeed, rather than seeking to stimulate economic activity, the aim of government
public health policy was to suppress it (and the social contact it involved) as a means of
controlling the spread of the virus until a vaccine was available. The focus of economic
policy at the height of the pandemic was to protect the incomes of households and the
survival of businesses while a significant part of the economy was closed. The closest
peacetime analogue for a shock of this nature to the UK economy was the more deadly
1918 flu pandemic. The economic impact of this earlier pandemic is, however, difficult to
disentangle from the effects of post-World War I demobilisation and its fiscal impact was
muted by the fact that both public health systems and the welfare state were in their infancy.

2.4 The coronavirus pandemic has affected not only the UK but nearly every country around the
world, bringing about the largest and most synchronised peacetime shock to the global
economy since the Great Depression of the 1930s. Global output fell by 3.3 per cent in
2020, far greater than the 0.1 per cent fall seen at the height of the global financial crisis in
2009. And regardless of how successfully they insulated themselves from the global spread
of the virus, few countries escaped its economic consequences, as shown in Chart 2.1.
Almost 90 per cent of economies suffered a decline in output last year, including every
advanced economy except Ireland and Taiwan.3 By contrast, 2009 saw output fall in only
half of all economies, while only a fifth suffered a decline in output in 1999, in the
aftermath of the Asian financial crisis and subsequent Russian debt default, and a third saw
falls in 1993, in the midst of a period that included the continuing effects of the Japanese
and Scandinavian financial crises, as well as Black Wednesday in the UK.

3
For Ireland, falls in domestic activity were more than offset by strong growth in activity in the multinational company sector (especially
pharmaceuticals and the ICT sector, which directly and indirectly benefitted from the pandemic despite the overall drop in global
demand). Quarterly National Accounts Quarter 4 2020,

Fiscal risks report 38

 Coronavirus pandemic

Chart 2.1: World GDP growth and the proportion of economies with falling output
6 120

5 100

4 80

3 60

2 40

Per cent
Per cent

1 20

0 0

-1 -20

-2 -40
Percentage of economies with output falls (right axis)
-3 -60
Real GDP growth (left axis)
-4 -80
1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 2016 2019
Source: IMF, OBR

2.5 It would be premature to draw definitive conclusions about the economic and fiscal
implications of the pandemic while the virus continues to circulate widely, economic activity
remains subject to public health restrictions, and extensive fiscal support remains in place.
Some economies that were spared the worst effects in the early phases of the pandemic
have suffered greatly in recent months. And the risks posed by new, vaccine-resistant
variants threaten even the vigorous economic recoveries seen in countries whose vaccination
programmes are furthest advanced.

2.6 However, as the largest peacetime shock to the UK economy and public finances in modern
memory, it is instructive to consider of the
pandemic in both historical and international context. By comparing it to the last major
economic shock, the 2008 financial crisis, we can identify both commonalities and contrasts
in the way in which the UK economy and public finances respond to different sources of
stress. By comparing experience during the pandemic with those of other countries,
we can better understand our sources of relative fiscal vulnerability and resilience.

Impact of the pandemic on the economy

Economic impact in historical context
2.7 The UK suffered its largest annual economic contraction in over three centuries as a result of
the coronavirus pandemic (Chart 2.2). The 9.8 per cent fall in real GDP in 2020 was over
twice as severe as the fall during the global financial crisis and over five times greater than
the average post-war recession. The economic shock resulted partly from voluntary social
distancing on the part of individuals trying to avoid contracting the virus and partly as a

39 Fiscal risks report

Coronavirus pandemic

result of government deliberately closing large sections of the economy.4 At the peak of the
first national lockdown in April 2020, economic output fell by 25 per cent relative to pre-
pandemic levels, while trips to retail and recreation areas fell by 78 per cent, transit levels
were down by around three quarters, and workplace attendance was down 70 per cent.

Chart 2.2: Annual GDP growth since 1701

25

20

15

10

5
Per cent

-5

-10

-15

-20
1700 1720 1740 1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980 2000 2020
Source: BoE, ONS, OBR

2.8 The coronavirus shock differed from previous UK recessions not only in its severity but also
in its degree of sectoral differentiation. While the financial crisis also saw declines in almost
every sector, the variation across sectors has been much greater during the pandemic, with
accommodation and food services falling by around 90 per cent while financial sector
output fell only 5 per cent. Over half of sectors saw declines of over 25 per cent in 2020,
while just one sector (mining and quarrying) saw a decline of that magnitude after the 2008
crash, with most sectors seeing falls of between 5 and 20 per cent. Output in the
accommodation and food services sector was still down 40 per cent in April 2021.

4
World Economic Outlook

Fiscal risks report 40

 Coronavirus pandemic

Chart 2.3: Sectoral peak-to-trough falls in 2008 and 2020

0
-10
Percentage change pre-shock level

-20
-30
-40
-50
-60
-70
-80
2020 fall
-90
2008 fall
-100

Source: ONS, OBR

2.9 Once businesses and consumers came to terms with the initial shock, the economy then
demonstrated a surprising degree of adaptability to each new round of public health
restrictions over the course of the pandemic. Following the sharp 25 per cent decline in
output under the first lockdown in Spring 2020, economic activity began to recover even
before public health restrictions were substantially eased. And subsequent lockdowns in
November 2020 and January 2021 saw smaller shortfalls relative to pre-pandemic activity
of 7 per cent and 9 per cent respectively, with the relationship between the stringency of
public health restrictions and economic output weakening over time (Chart 2.4). The rapid
IT-enabled shift to more people working remotely (with the proportion of workers working
from home rising from 27 per cent in 2019 to 47 per cent in April 2020) and more goods
and services being purchased online (with the share of total retail sales conducted online
rising from 20 per cent in January 2020 to 36 per cent in January 2021) greatly facilitated
this adaptation.5,6

5
Coronavirus and homeworking in the UK: April 2020, ONS, July 2020
6
Homeworking hours, rewards and opportunities in the UK: 2011 to 2020, ONS, April 2021.

41 Fiscal risks report

Coronavirus pandemic

Chart 2.4: The UK economy has adapted over successive lockdowns

0
Feb 20 Apr 21
Oct 20
Change in GDP relative to January 2020 (per cent)

Mar 21
-5
Sep 20 Feb 21
Mar 20 Aug 20
Dec 20
Jul 20 Nov 20 Jan 21
-10

-15 Jun 20

-20
No national lockdown
May 20
April 2020 lockdown
-25 Apr 20
November 2020 lockdown
January 2021 lockdown
-30
30 0 40 10 50 20 60 70 80 90 100
Change in stringency (relative to January 2020)
Source: Blavatnik School of Government, ONS, OBR

2.10 Economic activity also proved surprisingly resilient once public health restrictions were lifted.
At the corresponding stage of the global financial crisis, 15 months following the initial fall
in monthly output, output was still close to its lowest point of 7.0 per cent below the pre-
crisis level and did not regain that level for another 40 months. By contrast, output was only
4.0 per cent below its pre-pandemic level by April 2021, with 84 per cent of the drop in
output at the start of the pandemic having been made up. Our March 2021 forecast
assumed that activity would regain its pre-pandemic level within a further 12 months, more
than twice as fast as occurred following the 2008 financial crisis. Recoveries in output from
lockdowns have typically been sharper than we and various others predicted, driven by a
combination of stronger rebounds in both private and public sector activity.

Fiscal risks report 42

 Coronavirus pandemic

Chart 2.5: UK GDP recoveries: pandemic outturn versus forecast and financial crisis
105

100
Index (GDPat start of crisis = 100)

95

90

85
2008 financial crisis
80
Coronavirus pandemic
75
March 2021 forecast
70
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55
Months from start of crisis
Source: ONS, OBR

Economic impact in international context

2.11 While the pandemic led to the most synchronised global economic shock in modern history,
its impact differed in severity across countries. Countries that were able to contain the
spread of virus early have so far typically experienced shorter and shallower downturns and
faster recoveries, on average returning to pre-pandemic levels of activity at the start of
2021.7 Among the advanced economies in 2020, European countries generally suffered the
most economically, followed by those in North America and Japan, with other parts of Asia
and Australasia experiencing the least damage. But a striking feature of the pandemic has
been how much even those countries that did manage to contain the virus last year still
suffered economically, underscoring the spillovers that result from the open nature of
advanced economies and the consequences of global interconnectedness.

2.12 The UK experienced one of the deepest recessions among advanced economies last year,
with UK GDP falling by twice the advanced economy average in 2020.8 Only Spain, where
output contracted by 11 per cent, suffered a sharper fall among advanced economies (top
left panel of Chart 2.6).9 The relative severity of the downturn in the UK, which remains even
after allowing for cross-country methodological differences in the measurement of

7
See for example, SARS-CoV-2 elimination, not mitigation, creates best outcomes for health, the economy, and civil liberties, Barton et al,
The Lancet (2021), The 12-month stretch, Where the Government has delivered and where it has failed during the Covid-19 crisis,
Resolution Foundation, March 2021, COVID-19: Lockdowns, Fatality Rates and GDP Growth, König and Winkler, Intereconomics Vol 56
Number 1, 2021, Cross-country effects and policy responses to COVID-19 in 2020: The Nordic countries, Gordon et al, Economic Analysis
and Policy Volume 71, September 2021.
8
Box 2.4 of our March EFO looked in detail at international comparisons of the economic impact of the pandemic up to the third quarter
of 2020. On the current vintages of data, the UK experienced a larger fall in output in 2020 than most other major advanced economies,
even after adjusting for differences in the measurement of government output.
9
World Economic Outlook, IMF, April 2021.

43 Fiscal risks report

Coronavirus pandemic

government output, is likely to be a consequence of several factors.10 These include the fact
that the UK:

suffered high rates of coronavirus infections, hospitalisations and deaths among

advanced economies in 2020 (top right panel of Chart 2.6);

spent more time under stricter public health restrictions than most other advanced
economies (second only to Italy among major advanced economies) (bottom left); and

was more economically vulnerable to the pandemic by virtue of its relatively high share
of social consumption in output (second only to Spain among major advanced
economies) (bottom right).

Chart 2.6: Real GDP loss versus contributing factors: cross-country comparisons
2020 GDP change Deaths per million and GDP
0 0
2020 per cent change in GDP)

Korea
-2
-2
Australia
2020 per cent change in GDP

-4 New Zealand
Netherlands US
-4
-6 Japan Germany
Canada
-8 -6
Belgium

-10 -8 France
Italy
-12
-10 UK
Spain
-12
0 1000 500
1500 2000 2500
Deaths per million
Source: IMF, ONS Source: Blavatnik School of Government, IMF, ONS, OBR

Stringency and GDP Share of social consumption and GDP

0 0
Korea Korea
2020 per cent change in GDP

-2 -2
2020 per cent change in GDP

New Zealand Netherlands Australia Australia

New Zealand
US US Netherlands
-4 -4
Germany Japan Germany
Japan Canada
Canada -6
-6 Belgium
Belgium
-8 France
-8
France Italy
Italy
-10 UK
-10 UK
Spain
Spain -12
-12 10 15 20 25
30 40 50 60 70 Social consumption (per cent)
Average stringency in 2020 Note: The proportion of total household spending allocated to spending on
recreation and culture, and restauarants and hotels.
Source: Blavatnik School of Government, IMF, ONS, OBR
Source: OECD, IMF, ONS, OBR

10
For a more complete discussion of cross-country differences in the measurement of output, see Box 2.4
in our March 2021 Economic and fiscal outlook.

Fiscal risks report 44

 Coronavirus pandemic

Impact of the pandemic on the public finances

Fiscal impact in historical context
2.13 The pandemic also imparted the largest peacetime shock to . Public
sector net borrowing rose to 15.6 per cent of GDP in 2020-21,11 the largest deficit since
1944-45 and around one and a half times larger than the previous peacetime peak of 10.1
per cent of GDP reached in 2009-10 in the wake of the financial crisis. As was the case
after the two world wars but unlike after the financial crisis borrowing is forecast to fall
back quickly to near its pre-pandemic level within a few years. And the medium-term tax
rises and spending cuts the Government has announced are sufficient to eliminate all but a
£0.9 billion (0.03 per cent of GDP) current budget deficit in 2025-26. As a result of the
pandemic, public sector net debt is set to rise by over 20 per cent of GDP, but by
substantially less than the total increase in debt following the financial crisis, to peak at
109.7 per cent of GDP in 2023-24, its highest level since 1958-59.

Chart 2.7: Public sector net borrowing and net debt since 1900
PSNB PSND
30 300
March 2021 forecast March 2021 forecast
27 275
Outturn Outturn
24 250
21 225
18 200
Per cent of GDP
Per cent of GDP

15 175
12 150
9 125
6 100
3 75
0 50
-3 25
-6 0
1900-01 1925-26 1950-51 1975-76 2000-01 2025-26 1900-01 1925-26 1950-51 1975-76 2000-01 2025-26
Note: 2020-21 estimate includes our estimate of the costs of loan write -offs.
Source: ONS, OBR Source: ONS, OBR

2.14 The fiscal shock from the pandemic differed from previous recessions not only in its scale
and profile but also in its composition. The rise in borrowing in 2020-21 of 13 per cent of
GDP was driven almost entirely by an unprecedented discretionary increase in government
spending, rather than by the lower tax receipts we would usually expect to arise from
depressed economic activity (Chart 2.8). Higher spending on public services and support for
households and businesses was offset slightly by a fall in interest costs, thanks in large part
to the simultaneous expansion of the Bank of England quantitative easing programme and
cuts in Bank Rate. The rapid fall in borrowing from 2021-22 onwards primarily reflects the

11
This figure is based on the latest ONS estimate of PSNB in 2020-21 of £299.2 billion (released on 22 June), plus our March 2021 EFO
estimate that spending associated with loan guarantees will add £27.2 billion to borrowing in 2020-21. The ONS plans to reach its own
estimate of those write-offs to be incorporated in the official statistics later this year. Nominal GDP in 2020-21 is based on the ONS
estimate released on 22 June. The figure of 15.6 per cent of GDP is below the 16.9 per cent we estimated in our March 2021 EFO,
largely due to pandemic-related departmental spending coming in considerably lower than expected. In this section we use this estimate
of 2020-21 outturn when discussing the rise in borrowing due to the pandemic, but use our March 2021 forecast as it stood when
olicy
tightening in the medium term. This reflects both the provisional nature of the latest 2020-21 outturn and that some of the detail we need
to analyse the future path of borrowing is not yet available.

45 Fiscal risks report

Coronavirus pandemic

one-off nature of coronavirus rescue spending, which plans assume will

be fully withdrawn over 2021-22. We discuss the risks around this deficit reduction path
later in the chapter.

Chart 2.8: Pandemic borrowing was driven more by spending than receipts
1200
2008 financial 2020 coronavirus
crisis recession recession

1000

800
£ billion

600

400

Spending
200

Receipts
0
2007-08 2009-10 2011-12 2013-14 2015-16 2017-18 2019-20 2021-22 2023-24 2025-26
Source: OBR

Fiscal impact in international context

2.15 The UK saw the fourth largest increase in government borrowing as a per cent of GDP in
2020 among the 35 advanced economies (after Canada, Norway, and Singapore).12 The
outsized fiscal impact of the pandemic in the UK is attributable to four main factors
distinguishing it from other advanced economies:

First, the UK had a deeper recession than most other countries, for the reasons
highlighted above. This both raised the automatic element of the fiscal response,
increasing headline borrowing, and reduced the denominator (GDP).

Second, the UK was hit harder by the pandemic itself, which put greater pressure on
health services.13

Third, the UK entered the pandemic with relatively low levels of spare capacity in its
health system.14
beds, nurses, and physicians compared to its peers (Chart 2.10).15 Additional health-
related spending amounted to 5.3 per cent of GDP in the UK, around three times as
large as the 1.8 per cent of GDP average across advanced economies. Vaccine

12
Fiscal Monitor, IMF, April 2021.
13
Comparing G7 countries: are excess deaths an objective measure of pandemic performance?, Health Foundation, June 2021.
14
Health Foundation, November 2020.
15
OECD Health indicators, 2019.

Fiscal risks report 46

 Coronavirus pandemic

development and purchases, by contrast, accounted for a relatively small share of

additional spending in the UK and elsewhere.

Fourth, the UK also started off with a working-age welfare system that offered lower
income replacement for those facing reduced hours or falling out of work than systems
in other large advanced economies.16 An additional temporary income protection
system that also covered better-paid employees and the self-employed (in the form of
the CJRS and SEISS) was therefore created from scratch, whereas such support was
already provided for in some other countries.

Chart 2.9: Increases in government borrowing in selected advanced economies in

2020
12
Increase in borrowing (Per cent of GDP)

10

Source: IMF, OBR

16
The UK has lower replacement ratios for those falling out of work than other large advanced economies, both on average and for most
groups. However, the UK is close to the advanced economy average in terms of overall net social expenditure, and non-pensioner cash
benefits, as a proportion of GDP. After shocks, Financial resilience before and during the Covid-19 crisis, Resolution Foundation, April
2021; Net replacement rate in unemployment, OECD, data extracted on 29 Jun 2021; Social expenditure database, OECD, data
extracted on 29 June 2021; The shifting shape of social security, Charting the changing size and shape of the British welfare system,
Resolution Foundation, November 2019.

47 Fiscal risks report

Coronavirus pandemic

Chart 2.10: International comparisons of health system capacity

40
Top
35
Average
30
Bottom
25
United Kingdom
20

15

10

0
Acute care beds per Acute care bed Intensive care beds Physicians per 1,000 Nurses per 1,000 CT scanners per
1,000 vacancy rate per 100,000 100,000
Source: OECD, OBR

Fiscal policy response

Fiscal policy response in historical and international context
2.16 the pandemic was very large by both historical and
international standards:

Relative to the financial crisis, our March forecast estimated that the UK
rescue package cost 16.2 per cent of GDP over 2020-21 and 2021-22, which is
almost ten times the 1.7 per cent of GDP in fiscal support provided in 2008-09 and
2009-10 in response to the financial crisis.

Relative to other advanced economies, the IMF estimates that the policy
response to the pandemic was the third largest among 35 advanced economies in per
cent of GDP terms after the United States and New Zealand.17

2.17 The unprecedented scale of the fiscal policy response was partly a function of the limits on
what monetary policy could do given the nature of the pandemic shock and constraints on
conventional monetary policy instruments. Interest rates in the UK were already close to all-
time lows on the eve of the pandemic at 0.75 per cent but were reduced further to 0.1 per
cent in early 2020, while the amount of gilt purchases under quantitative easing was almost
doubled. Only fiscal policy could deliver the targeted support necessary during a pandemic,
focused on the households, businesses, and public services hit hardest by the pandemic and
associated public health restrictions. Broad-based demand stimulus provided by monetary

17
These figures use estimates from the IMF Fiscal Monitor, April 2021.

Fiscal risks report 48

 Coronavirus pandemic

policy could only ever have played a secondary role in these circumstances where economic
activity was being deliberately restrained by government policy.

2.18 Reflecting the proficiency with which it can be targeted, the fiscal policy measures deployed
in the UK and other advanced economies were heavily tilted toward spending rather than
revenue. The coronavirus rescue package was over 90 per cent spending and less than
10 per cent revenue. That contrasts with the more modest fiscal stimulus during the financial
crisis, which was about one quarter spending and three-quarters revenue (as the main rate
of VAT was temporarily cut). Across advanced economies, pandemic rescue packages were
similarly weighted toward spending as opposed to revenue (Chart 2.12).

2.19 spending than

other countries, with around one-third of total fiscal support accounted for by health-related
costs (Chart 2.11). This was more than twice the average proportion of health spending in
, which tended to be dominated by support to either
households, employment, or firms. Countries such as the US and Canada spent relatively
more on household support, partly reflecting their decision to channel support initially
through out-of-work benefits rather than through employment subsidies though of course
both approaches ultimately support household incomes. The low share of public works
spending in the UK partly reflects the large increases that were announced in the March
2020 Budget, on the eve of the pandemic hitting the UK, which are therefore not counted as
part of the rescue package in the UK.

Chart 2.11: Fiscal support by recipient for selected advanced economies

30
Other
Public works
25
Firms
Employment
20
Households
Per cent of GDP

Healthcare
15

10

0
Spain France Italy Germany Canada Japan United United Average
Kingdom States
Note: These figures and chart average come from IMF estimates of measures between January 2020 and April 2021 in 10 economies
for which detailed data are available.
Source: IMF, OBR

2.20 The policy response to the pandemic was also marked by extensive use of unconventional
fiscal instruments, which can expose governments to fiscal risks for many years after they
have been deployed. These included loans, guarantees, equity injections, and quasi-fiscal

49 Fiscal risks report

Coronavirus pandemic

support provided through state-owned development banks and other public corporations.18
Chart 2.12 shows that use of these unconventional instruments matched, and in some cases
exceeded, more conventional tax and spending measures. Such unconventional support was
used most extensively in Italy, with targeted government guarantees for both firms and
households, in Germany, through increased lending by its state-owned development bank,
and in Japan, through lending by publicly-owned financial institutions. The UK also made
relatively extensive use of government-guaranteed loans to support large, medium, and
small businesses in the form of the Coronavirus Large Business Interruption Loan Scheme,
Coronavirus Business Interruption Loan Scheme, and Bounce Back Loan Scheme. Total
exposure under these guarantees totalled 16.1
of unconventional fiscal support the fourth largest among 35 advanced economies.

Chart 2.12: Fiscal support by instrument for selected advanced economies

40
Conventional
Spending
35
Revenue
Unconventional
30
Liquidity
Guarantees
25
Per cent of GDP

Quasi-fiscal
20

15

10

0
Nether- France Spain Belgium Italy Germany Canada Japan United United Average
lands Kingdom States
Note: These figures and chart average come from IMF estimates of measures between January 2020 and April 2021 in 16 economies
for which detailed data are available.
Source: IMF, OBR

2.21 Comparisons of the scale of unconventional measures are based on total exposures, which
may not be a good guide to their ultimate fiscal cost. As discussed later in this chapter, that
depends on the terms of the guarantees and other instruments and how recipients of them
fare over the lifetime of the support, which in turn will determine the extent to which
guarantees are called, or loans and equity investments are written off (or written down). The
true direct fiscal cost of these interventions will therefore not be known with any certainty for
several years.

The impact of fiscal policy in supporting revenues

2.22 The provision of fiscal support on an unprecedented scale helped to avert much worse
consequences for private sector incomes during the pandemic, and almost certainly for

18
Quasi-fiscal support through state-owned entities is captured as -sector-wide fiscal
-country analysis.

Fiscal risks report 50

 Coronavirus pandemic

longer-term economic scarring beyond it, than would have resulted had the Government
not intervened. Estimating economic and fiscal costs of doing nothing would require an
exercise in counterfactual catastrophising that would stretch credulity. However, as discussed
above, one of the striking features of the coronavirus shock from a fiscal perspective has
been how much tax revenues have held up despite the dramatic contraction in output during
the pandemic (Chart 2.8).19 While overall, nominal GDP fell by 5.3 per cent in 2020-21,
receipts only fell by 4.1 per cent despite tax cuts that on their own would have left receipts
down 3.0 per cent and the dramatic falls in receipts from tax bases hit by public health
restrictions (with fuel duty down 24.2 per cent and air passenger duty down 90 per cent).
Much of the resilience of receipts is likely to be attributable to the extensive fiscal support
provided to protect household incomes and facilitate the survival of viable businesses.

Chart 2.13: Percentage change in different tax streams from 2019-20 to 2020-21
20

0
Percentage change on a year earlier

-20

-40

-60

Due to policy measures

-80

Due to underlying changes

-100
GDP Total Alcohol Council IT & NICs CT Tobacco VAT Fuel duty Business APD Other
receipts duties tax duty rates
Note: Uses 22 June 2021 Public Sector Finances release and 12 May 2021 first quarterly GDP estimate.
Source: ONS, OBR

2.23 As an illustration of the extent to which fiscal support measures also supported receipts, we
can compare the actual fall in receipts witnessed last year of £34 billion against a simple
baseline in which they fell in line with nominal GDP and the cost of tax cuts. That baseline
fall would have been £69 billion (with the 5.3 per cent fall in nominal GDP explaining £44
billion and tax cuts the remaining £25 billion). Tax receipts typically fall slightly faster than
nominal GDP in recessions due to the effects of fiscal drag going into reverse and sharper
falls in taxes linked to asset prices, which themselves vary more than one-for-one with GDP
so even this baseline could somewhat understate the relative strength of receipts last year.

19
How did COVID affect government revenues, spending, borrowing and debt?, IFS, June 2021

51 Fiscal risks report

Coronavirus pandemic

2.24 Chart 2.14 shows the sources of the £35 billion (1.7 per cent of GDP) outperformance of
receipts relative to this baseline:

Taxes on incomes and profits (income tax, National Insurance contributions, and
corporation taxes) held up much better than would have been expected given the fall
in nominal GDP. For personal income taxes, that reflects the fact that wages and
salaries grew by 1.4 per cent despite the fall in output thanks in particular to the £58
billion of support provided through the CJRS. For corporation tax, it reflects the support
for taxable profits relative to sales delivered by the £16.3 billion of grants and £10.4
billion of business rates relief that were provided in 2020-21. Overall, strength of the
PAYE income tax and NICs tax base and corporation tax combined explain three-
quarters of the receipts outperformance relative to the baseline (£26 billion).

Taxes on consumption (dominated by VAT) fell broadly in line with output. For VAT,
that reflected the main component of the tax base consumer spending falling more
sharply than nominal GDP (which on its own would have taken receipts down a further
£9 billion), but receipts holding up relative to that thanks to the almost fully offsetting
impact of VAT paid on higher government procurement, little of which was refundable,
and strength in other components of the tax base (such as the financial sector and
home improvements). The performance of other consumption taxes varied, with
alcohol duties particularly strong (up £1.6 billion relative to baseline).

Taxes on transport-related activity (fuel and air passenger duties) fell much more
sharply than GDP due to the effects of public health restrictions, but also account for a
much smaller share of total revenues in normal times.

Other receipts also outperformed the baseline. In part that reflects real-world
outperformance as with council tax, where the tax base is relatively fixed, or alcohol
duties but in part it reflects how some components of receipts are measured (for
example, much of public sector gross operating surplus simply equals depreciation,
which is linked to the public capital stock and is therefore invariant to GDP).

Fiscal risks report 52

 Coronavirus pandemic

Chart 2.14: Contributions to the strength of receipts in 2020-21

0

-10
Contributions to year-on-year change (£ billion)

-20

-30

-40

-50

-60

-70

-80

-90

-100
Nominal Fall due to Baseline Consumer Fuel duty PAYE IT & VAT vs Council Corporation Other Actual
GDP tax cuts fall spending and APD NICs tax consumer tax tax fall
related fall VAT tax base spending
base
ce the massive
Source: ONS, OBR increase in husehold saMedium-term

Post-pandemic fiscal plans in historical and international context

2.25 With the sharp rise in the deficit due almost entirely to temporary policy interventions, a
sharp fall in borrowing is forecast once the pandemic ends and this support is withdrawn.20
The fall in borrowing for the next five years would
be the largest and the fastest in peacetime. But there are risks to this planned reduction in
the deficit over the medium term, in particular from a potential legacy of additional
spending pressures that the pandemic and associated lockdowns could leave for public
services. In the face of such pressures,
around £14½ billion (0.6 per cent of GDP) a year from Departmental Expenditure Limits
(DELs) from 2022-23 onwards relative to its pre-pandemic plans, which contributes to the 2
per cent of GDP in medium-term fiscal consolidation relative to those plans that was
announced in the November Spending Review and the March 2021 Budget.

20
This section is based entirely on our March 2021 forecast, so does not reflect the lower initial ONS outturns for 2020-21.

53 Fiscal risks report

Coronavirus pandemic

Chart 2.15: The UK in historical context

0
Five-year cumulative change (per cent of GDP)

-5

-10

-15

Spending
-20
Receipts

PSNB
-25
World War I World War II Financial crisis Coronavirus pandemic
(1918-19 to 1923-24) (1945-46 to 1950-51) (2009-10 to 2014-15) (2020-21 to 2025-26)
Source: Bank of England, ONS, OBR

2.26 As the economy reopens and emergency fiscal support is withdrawn, government borrowing
is forecast to fall from a peacetime high of 16.9 per cent of GDP in 2020-21 to 2.8 per cent
of GDP in 2025-26. Of this 14 per cent of GDP in deficit reduction, 12 percentage points
comes from the unwinding of pandemic-related support to households, firms, and public
services, with the remainder explained by the recovery of the economy and the tax rises
announced in the March Budget, which together raise the tax burden to its highest level
since the late 1960s. Abstracting from pandemic-related spending, total public spending is
broadly flat as a share of GDP between 2020-21 and 2025-26, with modest falls in
working-age welfare spending offsetting further rises in investment spending in line with
pre-pandemic plans.

2.27 Viewed relative to pre-pandemic medium-term plans, 2 per cent of GDP of fiscal tightening
has been announced, reflecting both discretionary spending cuts and tax rises introduced
since the start of the pandemic. Of this, 60 per cent of the consolidation comes from taxes,
principally the increase in the corporation tax rate from 19 to 25 per cent alongside freezes
in the personal allowance and higher rate threshold for income tax, while the remaining 40
per cent comes from unspecified reductions in the envelope for departmental spending. This
compares with a post-2008 financial crisis consolidation in which the Coalition Government
planned for and delivered a 20/80 split between tax and spending, although over a
considerably longer timeframe than initially envisaged.21

21
unced in that Budget were split
precisely 20/80 tax and spending by 2014-15, while the overall discretionary consolidation including the measures inherited from the
outgoing Labour Government were split 23/77 tax and spending by 2015-16. The IFS estimates that in outturn spending contributed
between 80 and 90 per cent of the total post-financial crisis consolidation (Fiscal response to the crisis, IFS).

Fiscal risks report 54

 Coronavirus pandemic

2.28 The net effect of the sustained increases in departmental spending planned in the March
2020 Budget and the cuts to those totals announced since then is to leave both total and
departmental spending higher in the medium term than they were in 2019-20 in cash
terms, real terms and as a share of GDP. Table 2.1 breaks down the rise in spending as a
share of GDP between 2019-20 and 2024-25 (the final year of our March 2020 forecast)
into contributions from: pre-pandemic spending plans and forecasts; the effect of nominal
GDP being weaker in the medium term; and the effect of changes in cash spending since
the March 2020 Budget. It shows that between 2019-20 and 2024-25:

Total managed expenditure (TME) rises by 2.1 per cent of GDP, thanks entirely to
higher departmental resource and capital spending. This increase in TME is 1.2 per
cent of GDP larger than we forecast in March 2020. This difference is more than
explained by weaker nominal GDP (adding 1.6 per cent of GDP), partly offset by the
£12 billion downward revision to cash spending in 2024-25 (subtracting 0.4 per cent).

Departmental resource spending (RDEL) rises by 1.0 per cent of GDP, 0.2 per cent of
GDP less than it did based on March 2020 Budget plans. On unchanged cash totals,
the weaker outlook for nominal GDP would have added 0.8 per cent of GDP to the
rise in spending over the medium term, but the £16 billion cut to RDEL totals in 2024-
25 announced since the March 2020 Budget offsets most of that (subtracting 0.6 per
cent of GDP when combined with the effect of spending in 2019-20 having been
revised up).

Departmental capital spending (CDEL) rises by 1.1 per cent of GDP, 0.2 per cent of
GDP more than it did on March 2020 Budget plans. This upward revision is split
equally between weaker nominal GDP and modestly faster growth in cash spending
(as unchanged 2024-25 plans are compared to downwardly revised 2019-20
outturn).

Other spending is flat. Our March 2020 forecast predicted a significant fall of 1.2 per
cent of GDP, but that has been lost to the effects of weaker nominal GDP (explaining
0.8 per cent of GDP) and higher cash spending (explaining 0.3 per cent of GDP). The
£4.1 billion upward revision to annually managed expenditure in 2024-25 is more
than explained by a £5.1 billion upward revision to welfare spending.

Table 2.1: Change in spending between 2019-20 and 2024-25

Per cent of GDP
Difference from Effect of
March 2021 March 2020 of Effect of lower changes in cash
forecast forecast which nominal GDP spending
Total managed expenditure 2.1 1.2 1.6 -0.4
of which
Departmental resource spending 1.0 -0.2 0.6 -0.8
Departmental capital spending 1.1 0.2 0.1 0.1
Other spending 0.0 1.2 0.8 0.3

55 Fiscal risks report

Coronavirus pandemic

2.29 The post-pandemic path of public expenditure and its allocation between competing
pressures and priorities will be the subject of the 2021 Spending Review, whose conclusions
are expected in the autumn. Ahead of those decisions, this section considers the potential
legacy of direct fiscal pressures that the pandemic could leave behind. To the extent that
these pressures are accommodated by increasing the total level of spending, this constitutes
a risk to the borrowing outlook and the
budget and getting underlying debt to fall as a share of GDP. To the extent that they are
accommodated within the spending envelope inherited from the March 2021 Budget, they
would imply reductions in the r
budgets are not covered by a pre-existing commitment to spend a particular sum of money.

Unwinding the pandemic rescue package

2.30 The Government plans for rapidly shrinking the deficit over the next five years depend
crucially on ending pandemic-related support to individuals, businesses and public services
by the end of the 2021-22 fiscal year. Its ability to do so clearly depends on the future
course of the pandemic, progress in lifting of remaining public health restrictions, and
prospects for a full recovery in different sectors of the economy. The emergence of the delta
variant of the virus, and the resulting rise case numbers and hospitalisations, have already
prompted the lifting of the final set of public health restrictions in England to be postponed
from 21 June to 19 July.

2.31 However, even if the reopening proceeds as now planned in July, there remain significant
risks associated with the winding down of support to firms and individuals. Those associated
with guaranteed loans to business and support to public services (principally health,
education, and transport) are discussed below. Risks associated with the unwinding of the
CJRS (furlough) scheme are considered alongside longer-term prospects for the labour
market in the next section. The planned withdrawal of the temporary £1,000 a year uplift to
the standard allowance in universal credit from October, which will reduce the cash incomes
of millions of families when it takes effect, is noted as welfare spending risk in Chapter 5.

Government-guaranteed loans
2.32 coronavirus-related guarantees on business loans present a material
source of fiscal risk over the medium term. Through a variety of schemes, the Government
has provided a mix of full and partial guarantees against potential losses incurred by
creditors worth up to £69 billion (3.1 per cent of GDP).22 These include:

£46.5 billion of exposure to potential losses through the Bounce Back Loan Scheme
(BBLS) for small businesses, which provides full compensation for losses on loans worth
between £2,000 and £50,000;

£18.6 billion through the Coronavirus Business Interruption Loan Scheme (CBILS) for
small and medium-sized organisations, which provides 80 per cent indemnification for
losses on loans worth between £50,000 and £5 million; and

22
HM Treasury coronavirus (COVID-19) business loan scheme statistics.

Fiscal risks report 56

 Coronavirus pandemic

£4.2 billion through the Coronavirus Large Business Interruption Loan Scheme
(CLBILS) for medium-sized and large businesses, which provides 80 per cent
indemnification for losses on loans up to £200 million.

Chart 2.16: Loans issued under the pandemic-related loan guarantee schemes
80
BBLS
70
CLBILS
60
Volume of lending (£ billion)

CBILS
50

40

30

20

10

0
10 May 31 May 28 Jun 26 Jul 16 Aug 20 Sep 18 Oct 15 Nov 13 Dec 24 Jan 21 Feb 21 Mar
2020 2021
Source: HMT, OBR

2.33 Through these three schemes, the Government guaranteed nearly half of all lending to
small and medium-sized businesses in 2020-21. This government-supported lending helped
to keep businesses afloat and avoid the kind of credit crunch that occurred during the
financial crisis. This is a different approach than was witnessed during the financial crisis. At
that time, the Government intervened to prevent the collapse of financial institutions
themselves, effectively providing compensation after the fact for losses that had already
crystallised. In this case, the Government acted early, guaranteeing individual lending
exposure and effectively providing insurance on losses before the fact.

2.34 As such, the G

significant cash costs in the event of firms defaulting on the underlying loans. For the largest
scheme, the BBLS, arrears have so far not arisen because none of these loans have fallen
due for repayments. Our forecasts assume future default rates on the loans that try to reflect
the inherent riskiness of each of the instruments. As shown in Chart 2.17, the expected cost
borne by the public sector is arrived at after taking off cash recovered from the borrower
and losses covered by the lender. These expected fiscal costs are reflected in the measure of
public sector net borrowing in the year the guarantees were extended (2020-21).

2.35 For CLBILS, aimed at medium and large businesses, which are more resilient to negative
shocks, we assumed only 10 per cent of guarantees would be called. For CBILS, aimed at
smaller businesses, we assumed 17.5 per cent. For both schemes, lenders bear a fifth of the
associated costs. But for BBLS, we assumed 45 per cent of guarantees would be called,

57 Fiscal risks report

Coronavirus pandemic

reflecting the greater riskiness of these borrowers with smaller businesses having a higher
likelihood of not repaying their loans and in the nature of the guarantee scheme with
BBLS guaranteeing all the amounts loaned out, whereas for CLBILS and CBILS only partial
guarantees were issued. Our latest estimated cost for lifetime claims on these three
schemes, published in our March 2021 EFO, is £26.1 billion.

Chart 2.17: Loans approved and expected fiscal costs of loan guarantees
100

90

80

70

60

50
£ billion

40

30

20

10

0
Loans approved Repaid loans Cash recovered from Losses covered Fiscal cost
by lender defaulting loans by lender
Note: figures based on our March 2021 forecast for loans approved and total losses, with illustrative further
breakdown based on assumptions in our 2020 Fiscal sustainability report.
Source: HMT, OBR

2.36 The ONS determined that expected losses on these schemes should be recorded upfront in
the public finances when the loans covered by the guarantees were provided. In our
forecasts and scenarios over the past year, we have included our best estimate of the size of
those expected losses in our PSNB figures. The ONS has not yet included them in the public
sector finances statistics it publishes every month. We have estimated these costs based on
analysis of similar past loan schemes, but the ONS will come to a view once the British
estimates compliant with the financial reporting requirements become
available.

2.37 There is, therefore, considerable uncertainty around the £26.1 billion expected loss on these
schemes reflected in our estimate of 2020-21 borrowing. The financial health of the
businesses that have taken out loans will depend on how the economy recovers as well as
risks at the individual company level. There is little evidence to date that would allow us to
gauge whether the probability of default implicit in our expected loss assumptions is too
high or low given the forbearance measures still in place including the recent extension of
protection for commercial tenants in rent arrears. As of May 2021, company insolvencies
were still down 25 per cent on May 2019 (pre-pandemic) and were up only 7 per cent from
the very subdued levels of May 2020. In addition to uncertainty around the extent of
company failures, it is not yet known the extent to which loans were drawn down
fraudulently by those taking advantage of the generous support on offer.

Fiscal risks report 58

 Coronavirus pandemic

2.38 Future policy changes also pose a risk. The terms of BBLS loans have already been relaxed
even before any repayments had to be made, when the Treasury
s businesses borrowing under the scheme to extend the term
of the loan from six to ten years; to make interest-only payments for six months (up to three
times); and to take a full repayment holiday of six months. And while BBLS, CBILS and
CLBILS were only in place for 2020-21, the Treasury announced a successor scheme the
Recovery Loan Scheme in the March 2021 Budget. These actions point to the risk of
further forbearance on existing loans and further extensions or successor schemes in future
as repayments start to come due.

2.39 The Government balance sheet is also exposed to risks around its growing portfolio of
equity investments. This has been formalised through the Future Fund, which allows
companies to apply for equity-convertible loans of up to £5 million. The Treasury
progressively increased the amount it was willing to allocate to the convertible loans under
the scheme, from an initial £250 million to £1.1 billion for all applications approved by 21
February.23 Our forecast assumes a 30 per cent loss rate over three years, but even on that
basis write-off costs would be small relative to those associated with the larger loan
guarantee scheme willingness to
make active use of its balance sheet to support non-financial corporations, mirroring the
approach taken with financial institutions in the wake of 2008-09 crisis. Estimates by
Beauhurst24 point to at least 25 (out of 1,236) loans to companies having already been
converted to equity, leaving the Government with stakes in varied companies such as a low-
flush toilet maker, a broadband provider, a reusable packaging producer and a satellite
company. We asked the Treasury to provide us with the latest positio HM
Government has not published data on the amount of conversions awarded by the Future
Fund, but has regularly published data including value of convertible loans awarded and the
diversity statistics of the funding from the scheme .

Post-pandemic pressures on public services

2.40 Another key source of direct pressure on the public finances comes from the legacy the
pandemic leaves behind for a range of public services. The very large sums allocated to
fight the virus mean that departmental resource spending (RDEL) was expected to have risen
by up to 39 per cent or £124 billion in 2020-21 in our March 2021 forecast. A smaller £56
billion (equivalent to a 15 per cent increase on pre-pandemic RDEL plans) has been added
to fund virus-related activities in 2021-22, but no provision for virus-related spending has
been made in 2022-23 and beyond. Instead, core spending totals from 2022-23 onwards
were cut by around £14½ billion a year in the November 2020 Spending Review and the
March 2021 Budget relative to plans set out in the March 2020 Budget. At the same time,
as shown in Table 2.1 above, both total spending and departmental spending is still
expected to be higher in 2024-25 than in the year before the pandemic in 2019-20.

23
Applications closed on 31 January 2021, but processing is still ongoing for those submitted before closure.
24
UK government becomes shareholder in toilet maker

59 Fiscal risks report

Coronavirus pandemic

2.41 The scale of potential post-pandemic departmental spending pressures, which amount to
around £12 billion next year and decline to around £9 billion after three years, are
considered in more detail in this section. These unfunded pressures are comprised of
approximately £7 billion a year in pressures on the health service, £1¼ billion a year in
education, and declining amounts that average £2 billion a year in transport (Chart 2.18).
These figures are subject to varying degrees of uncertainty and represent only a subset of
the universe of pandemic legacy spending risks for Government departments, but we
consider them to be reasonable estimates of some of the larger potential post-pandemic
pressures on DEL. They largely draw on external sources in the absence of detailed
estimates from the Government. And
we have not attempted a comprehensive assessment of potential savings that might stem
from the pandemic. Nor has the Government proposed any.
decision to suspend multi-year budget planning and revert to annual spending rounds for
most departments in recent years, whether and how the Government chooses to respond to
these pressures is not yet known.

Chart 2.18: Illustrative estimates for selected pandemic-related pressures on

departmental resource spending
14
Transport
Education
Health: Test & Trace, and vaccinations
12
Health: Pandemic legacy pressures
Health: NHS backlog
10

8
£ billion

0
2022-23 2023-24 2024-25
Note: Assumptions and sources underpinning estimates of individual pressures can be found in Chapter 2.
Source: OBR

Health
2.42 Pandemic-related pressures on health spending could amount to £7 billion a year on
average over the next three years, with pressures likely to be greatest in the near term. The
larger sources of potential pressure include: maintaining a standing capacity for test and
trace and vaccinations; addressing the backlog of elective treatments built up during the
pandemic; and the implications for NHS productivity of building in greater resilience and
the greater capacity for infection control than was allowed for in pre-pandemic plans.

Fiscal risks report 60

 Coronavirus pandemic

2.43 The most direct virus-related risks to the associated

with coronavirus itself. The Department of Health and So -
virus budget in 2021-22 was set at £147.1 billion in the 2020 Spending Review. While an
additional £63 billion was added in 2020-21 and £29 billion was added in 2021-22, there
have been no additions to its pre-pandemic multi-year settlement thereafter. There are
several potential additional demand pressures as a direct result of the pandemic, including:

Controlling the virus as it continues to circulate. So long as the virus remains prevalent
in the UK, there are likely to be ongoing costs from NHS Test and Trace, for which the
Treasury has allocated £15 billion in 2021-22. With vaccinations providing some
degree of protection against infection, it is very unlikely that ongoing costs would be
anywhere near as high as they have been to date and they might also be expected to
become more concentrated in winters rather than spread throughout the year. If we
therefore assume that Test and Trace spending will be required for three months a
year, but at a monthly cost that halves each year (reflecting some combination of fewer
tests being administered and/or unit costs falling), this might cost £2 billion next year,
falling to £1 billion in 2023-24 and £½ a billion in 2024-25.

Ongoing costs from vaccinations and revaccinations. The Government noted in its
February vaccinations
regular part of managing COVID-19 a revaccination
campaign, which is likely to run later this year in autumn or winter... on the basis that
[the NHS] will need to run COVID-19 and seasonal flu vaccination campaigns in
parallel. 25 The relatively low unit cost of purchasing and administering vaccines
(around £10 per dose)26 means that providing two to each adult
in the UK (at a take-up rate of 95 per cent) would cost just over £1 billion a year.

Greater-than- . Around 1million

individuals in the latest ONS survey self-reported long-Covid symptoms in May 2021,
with 376,000 reporting symptoms more than a year after they had had the virus.27 A
more recent DHSC study suggests that the number of people reporting long-Covid
symptoms was higher still at over 2 million.28 Whether and how these cases might
subsequently translate into any additional costs for the health (or welfare) system is
unknown at this stage. We therefore do not include a long-Covid-related estimate in
our summary of pandemic-related pressures on health spending.

The consequences for mental health arising from the pandemic and the lockdowns.
The Health Foundation REAL Centre projected that referrals to dedicated mental health
services for adults and children could increase by an average of 11 per cent in the
aftermath of the pandemic. Absorbing this increase in caseload could cost the health
service £1.1 billion next year, rising to £1.4 billion by 2024-25.29 There may also

25
NHSCOVID-19 vaccine deployment, NHS England and NHS Improvement Board meetings, March 2021.
26
COVID-19: Planning for the vaccine (part 1) Inquiry, NAO, 2021.
27
Prevalence of ongoing symptoms following coronavirus (COVID-19) infection in the UK: 4 June 2021, ONS.
28
New research shows 2 million people may have had long COVID, DHSC, June 2021.
29
, Health Foundation REAL Centre, November 2020.

61 Fiscal risks report

Coronavirus pandemic

knock-on impacts on future welfare spending (which we do not capture) if the doubling
in self-reported prevalence of depression fed through to higher numbers of GP
diagnoses as services return to normality and subsequent inflows to universal credit
and disability benefits.30

2.44 In addition to these direct demands on the health service, the Government stated in the
committed to building resilience for any future pandemics, both
domestically and on the international stage. greater spare
capacity in the health service so that it is more resilient to sudden surges in demand of the
type experienced over the past year. As discussed above, in comparison to other advanced
economies, the UK entered the pandemic with relatively low per capita numbers of critical
care beds and relatively high levels of bed occupancy. The NHS estate might also need to
be reconfigured so that managing large numbers of infectious patients and segregating
them from the non-infected population does not routinely disrupt other treatments. (Health
sector output fell by 15 per cent in 2020 as hospitals redesigned infection prevention and
control to address the new coronavirus risks.31) The Health Foundation notes that continued
social distancing and infection control measures could reduce NHS productivity relative to
pre-crisis assumptions, calculating that every percentage point of productivity lost could
generate £1.4 billion to £1.7 billion a year of spending pressure.32 If NHS productivity were
to suffer a hit of 1.2 per cent in line with our economy-wide TFP scarring assumption this
would imply around £1.8 billion a year in additional cost pressures.

2.45 In addition to these virus-related pressures, there may be costs associated with clearing the
backlog of non-virus-related treatments in the NHS. Between April 2020 and May 2021
there were 3.5 million fewer elective procedures and over 22 million fewer outpatient
attendances in England than over the same period in 2019-20.33 At least some of those
people not seen last year will need treatment eventually, which can be expected to add to
the 5.1 million already on a waiting list for NHS care. Delayed treatment might also mean
that their health is now worse and that the cost of treatment will be higher. Waiting times
have already risen: the latest figures show that 385,000 people have been on NHS waiting
lists for more than a year, compared to just over 1,500 before the pandemic.34 The Health
Foundation estimated that tackling the backlog of demand for elective care and restoring
waiting times to pre-pandemic standards would cost £1.9 billion a year over three years
(while also warning that the level of increased activity required to do so might not be
achievable due to staffing constraints).35

30
Are we facing a mental health pandemic?, ONS, May 2021.
31
UK Economic Accounts: main aggregates, ONS, 31 March 2021.
32
Spending Review 2020: Managing uncertainty, COVID-19 and the NHS long term plan, Health Foundation, November 2020.
33
Pressure points in the NHS, British Medical Association, June 2021.
34
Consultant-led Referral to Treatment Waiting Times Data 2021-22, NHS England, June 2021.
35
clearing the backlog was made prior to the third national lockdown. A more recent estimate
factoring in additional elective treatments postponed this year could therefore be higher still. Its estimate was based on the then
million fewer patient referrals compared with the same months in 2019. Assuming that 75 per cent
treatment and are referred by the end of 2020/21 [means] the waiting list would grow to 9.7 million by 2023. Clearing this backlog over 3
years, while treating the expected normal growth in referrals by 2023/24, would require treating 1.5 million more patients a year beyond
the long-

Fiscal risks report 62

 Coronavirus pandemic

Chart 2.19: NHS waiting lists

6

Waiting more than a year

5

Waiting less than a year

4
Million

0
Aug 07 Aug 08 Aug 09 Aug 10 Aug 11 Aug 12 Aug 13 Aug 14 Aug 15 Aug 16 Aug 17 Aug 18 Aug 19 Aug 20
Source: NHS England

Education
2.46 The closure of schools and other education establishments for extended periods over the
past year has significantly reduced the number of teaching hours received by the current
cohort of school-aged children. The studies suggesting
the total loss in face-to-face learning could amount to around half a school year . The Prime
no child will be left behind as a result of the pandemic the
develop a long-term plan to make sure pupils have the chance to
make up their learning over the course of this Parliament Sir Kevan Collins was appointed
as an Education Recovery Commissioner in February 2021 to oversee a comprehensive
programme of recovery aimed at young people who have lost out on learning due to the
pandemic He resigned in June 2021 and no such programme has been forthcoming.

2.47 At the time of our March forecast, the Government had already committed £1.7 billion to
catch-up education spending. Since then, it has announced a further £1.4 billion to fund
extra tuition for some pupils.36 The Prime Minister subsequently described this as being just
for starters 37 and the Government has said that education recovery will be reviewed further
at th Spending Review. The extent of further pressures this could pose are highly
uncertain given the different types of intervention that could be pursued. In addition to extra
tuition the Education Policy Institute estimated that allocating £3.2 billion over the next three
38
years for extended school hours would be sufficient to recover two
At a cost of £600 per pupil per year, this would imply a pressure of around £1 billion a year
over the next three years in addition to the amounts already announced.

36
Huge expansion of tutoring in next step of education recovery, Department for Education, June 2021.
37
Questions, 9 June 2021.
38
Education recovery and resilience in England, Phase one report, Education Policy Institute, May 2021.

63 Fiscal risks report

Coronavirus pandemic

Transport
2.48 The pandemic has also significantly disrupted domestic and international transport and
generated calls for substantial and lasting fiscal support to the sector. The Government has
already intervened in the past year with direct support to the railways and to Transport for
London at a cost of £12.8 billion in 2020-21. The National Infrastructure Commission has
presented a range of possible scenarios for the enduring impact of the pandemic on public
transport numbers out to 2055.39 These scenario involving
up to 10 per cent fewer public transport trips up to scenario in which
they were 25 per cent lower. As of 28 June, use of the railways remained down 55 per cent
relative to pre-pandemic levels and use of London Underground was still down 54 per
cent.40 Given rail income of around £11.6 billion in 2019-20,41 and accounting for
inflation, assuming a 25 per cent shortfall in 2022-23
scenario) that eases to 10 per cent by 2024-25 (in line
scenario) would imply revenue losses and thus a spending pressure of £3.0 billion in 2022-
23 that would diminish to £1.2 billion a year by 2024-25.

2.49 Public and private providers have typically relied on relatively better-off commuters with
limited choice travelling at peak hours to pay the bulk of fares while in effect subsidising the
travel of off-peak travellers. Transit on trains, buses and urban metros fell across the world,
but it is striking that usage in the UK has stayed lower for longer than other comparator
countries, and is currently more than half as much again below pre-pandemic levels as in
these countries. The shift to working from home for sections of the economy could threaten
this decades-old funding model first through lower traffic in total, but also through
reducing the concentration of passenger numbers at particular times of the day that allows
providers to charge higher prices during predictable periods of peak demand.

39
Behaviour change and infrastructure beyond Covid-19, National Infrastructure Commission, May 2021.
40
Transport use during the coronavirus (COVID-19) pandemic, Department for Transport, June 2021
41
Rail Industry Finance (UK) 2019-20, Office of Rail and Road, November 2020.

Fiscal risks report 64

 Coronavirus pandemic

Chart 2.20: Changes in public transport mobility during the pandemic

40
Other advanced economies
United Kingdom
20
Advanced economy average
Index (0 = pre-pandemic level)

-20

-40

-60

-80

-100
Feb 20 Apr 20 Jun 20 Aug 20 Oct 20 Dec 20 Feb 21 Apr 21 Jun 21
Source: Our world in data, accessed on 25 June 2021

Other potential pressures on departmental spending

2.50 The estimates presented above cover some of the larger pandemic-related pressures, but
they do not represent a comprehensive assessment of such pressures or indeed potential
savings, for example from greater delivery of public services online. Nor do they cover other
pressures that are less directly related to the pandemic. These issues would include:

Any pressures from long Covid , which as noted above are currently unknown.

Prospective reforms to adult social care, which have been under consideration by
successive governments for the past decade, and where
2019 manifesto stating urgently seek a cross-party consensus in order to
bring forward the necessary proposal and legislation for long-term reform .

The cost of addressing pandemic-related backlogs in the justice system on the Ministry
RDEL budget, which was £9.3 billion in 2021-22.

The cost of restoring Official Development Assistance spending to the legislated target
of 0.7 per cent of GNI from the 0.5 per cent it was temporarily reduced to in Spending
Review 2020. 0.2 per cent of GNI is equal to £4.7 billion a year in 2022-23, rising to
£5.2 billion a year in 2025-26 thanks to continuing economic growth.

65 Fiscal risks report

Coronavirus pandemic

Long-term economic legacy of the pandemic

2.51 Beyond any direct medium-term pressures, the longer-term fiscal risks associated with the
pandemic will depend on its lasting impact on potential GDP and demographic trends. As
set out in our last two EFOs, we have so far assumed in our central forecast a
effect (defined as the shortfall in potential output relative to its pre-pandemic trajectory at
the five-year forecast horizon) of 3 per cent. But given the extreme uncertainty about the size
of this effect, we also presented two alternative scenarios: an upside path with no long-term
scarring on output; and a downside path assuming long-term scarring of 6 per cent.

2.52 This range of scarring estimates was broadly in line with external estimates for the UK
economy and official forecasts for other European countries. The range was based on top-
down judgement rather than precise bottom-up modelling and did not presume a
mechanical connection to specific near-term policies or developments. That
notwithstanding, in our November EFO, we presented a putative decomposition of our 3 per
cent central scarring assumption:

Lower investment during the pandemic and subsequent recovery lessening the amount

accounted for 0.8 percentage points.

Lower total factor productivity (TFP) reflecting reduced investment in R&D during the
pandemic, together with the assumption that the ongoing presence of the virus would
require some businesses to adopt less efficient ways of operating (such as more
distancing within workplaces). Higher business debt and firm failures should also
weigh on future innovation. This accounted for 1.2 percentage points.42

Lower labour supply, accounting for 1 percentage point. Within this, half was down to
lower participation, reflecting the longer-run health consequences for some of those
contracting the virus and a decision by some older workers to retire earlier. The
remainder was split roughly equally between modestly higher unemployment (as
workers moved across jobs, sectors and occupations) and a smaller population (as a
result of lower net migration). Average hours worked per person was assumed to
return to their pre-pandemic trajectory, so did not contribute to labour market scarring.

2.53 Over the longer term, the loss of face-to-face education by students would also be expected
to have an adverse impact on their subsequent productivity and be reflected in lower lifetime
earnings.43 We did not consider this channel for our medium-term forecast as the effect
would mostly occur beyond our forecast horizon.

42
In reality, some of the TFP shortfall would also reflect capital scrapping as a result of business failures or faster depreciation of the
remaining capital stock due to the adoption of new and less efficient modes of operation as result of the virus. But effects of this sort
are unlikely to be picked up in the official capital stock statistics, so would instead show up in measures of TFP.
43
Costs of lost schooling could amount to hundreds of billions in the long-run, IFS, February 2021.

Fiscal risks report 66

 Coronavirus pandemic

Recent data and analysis

2.54 Data and analysis released since we made our original judgements paint a mixed and
incomplete picture on scarring:

The ONS has revised up its estimates of business investment. At the time of our
November 2020 forecast, business investment in the second quarter of 2020 was
estimated to have been 27 per cent below its level in the fourth quarter of 2019 before
recovering to be 20 per cent below in the third quarter (Chart 2.21). These figures
have since been revised up to 23 per cent and 13 per cent respectively. Data for the
fourth quarter of 2020 and the first quarter of 2021 also point to smaller shortfalls
than assumed in our November 2020 forecast at 7 and 17 per cent rather than 26
and 22 per cent respectively. This suggests the impact of the capital shallowing
channel might be less than we originally thought.

Chart 2.21: Business investment during the pandemic

105

100

95

90
2019Q4 = 100

85

80

75
March 2020 outturn and forecast
November 2020 outturn and forecast
70
March 2021 outturn and forecast
Latest data
65
2019Q4 2020Q1 2020Q2 2020Q3 2020Q4 2021Q1
Source: ONS, OBR

There is little new information regarding the impact on TFP, but external analysis of the
that was published between
our November and March forecasts suggests that the pandemic could reduce private
sector TFP by around 1 per cent in the medium term.44 While the successful vaccine
rollout has facilitated a faster recovery in output in recent months than we expected
and might be consistent with a better financial position for firms, the additional
lockdown at the start of this year will have led to a further deterioration for some
businesses and might lead to the loss of firm-specific knowledge from more firm
failures, while additional debt incurred might weigh on future innovation.

44
N Bloom et al. The Impact of Covid-19 on Productivity, NBER Working Paper 28233, December 2020.

67 Fiscal risks report

Coronavirus pandemic

Analysis of labour market data, discussed in Chapter 2 of our March 2021 EFO,
suggested that the working-age population may be substantially smaller than
incorporated into the official statistics. This would be the result of significant numbers
of foreign-born nationals returning home during the pandemic and lower levels of
immigration than pre-pandemic projections assumed. The ONS has subsequently
released new analysis and has set out plans to improve the evidence base in this
area.45 Initial experimental modelling by the ONS suggests that net migration fell
during the initial phase of the pandemic, to a net outflow of around 67,000 between
March and June 2020.46 ONS analysis of -time information (RTI) from the
PAYE tax system suggests that the population in the fourth quarter of 2020 could be
around ½ per cent smaller than currently incorporated into labour market data.47 This
analysis implies that the impact of the population scarring channel might be greater
than we originally expected.

The official Labour Force Survey (LFS) suggests that the unemployment rate has been
lower than we expected at 4.8 per cent in the first quarter of 2021 compared to 5.1
per cent in our November forecast. RTI data are consistent with a somewhat higher
unemployment rate of around 5.5 per cent (with the gap relative to the LFS having
narrowed slightly as the number of payrolled employees picked up in April and May).
There were also still around 2.6 million people on furlough in May (about 8 per cent
of the labour force), some of whom are likely to flow into unemployment over the
coming months.

The participation rate was 63.4 per cent in the first quarter of 2021 compared to 63.7
per cent in our November forecast, and is down 0.8 percentage points relative to the
first quarter of 2020. The pandemic has so far had a larger impact on labour market
participation among both older workers and younger workers relative to those in
middle of their working lives (Chart 2.22)48 While the latest statistics show that the
change in participation levels is mainly driven by the young, a significant number of
people over 65 have also left the labour market, halting the recent trend of increasing
participation for this age group. This could be indicative of older workers taking earlier
retirement following the pandemic, which would lower overall participation relative to
pre-pandemic assumptions. Relatively few forecasters have included a participation
channel in assumptions about medium-term scarring.

45
ONS, Population and migration statistics system transformation overview, April 2021 and June 2021.
46
ONS, Using statistical modelling to estimate UK international migration, April 2021.
47
ONS, Labour Force Survey weighting methodology, May 2021.
48
N. Comminetti, U-Shaped Crisis, April 2021.

Fiscal risks report 68

 Coronavirus pandemic

Chart 2.22: Contributions to change in participation between 2020Q1 and 2021Q1

0.2

0.0

-0.2
Percentage points

-0.4

-0.6

-0.8

-1.0
16-17 18-24 25-34 35-49 50-64 65+ All adults
Age group
Source: ONS, OBR

Our putative breakdown did not include an average hours effect, but the pandemic
could have lasting consequences on working patterns. It has accelerated the
movement towards working from home, with the proportion of the workforce who did
some work at home rising to 35.9 per cent in 2020, up 9.4 percentage points from
2019. Recent data from the ONS BICs survey suggest around a quarter of businesses
plan to continue increased home working. The consequence of this for average hours
is presently unclear. One the one hand, full-time workers who mainly work from home
tend to work more hours on average than those who never work from home. But on
the other, those who mainly work from home are more likely to work part time than
those who never work from home.49

udgements
2.55 In its latest World Economic Outlook released in April 2021, the IMF estimated that the
pandemic would lower world output in 2024 by around 3 per cent relative to their pre-
pandemic forecast, albeit with significant variation across countries. This is significantly less
than the IMF estimates of an average 6 per cent loss for past pandemics and epidemics,
and the almost 10 per cent loss following the financial crisis. While the shock to global
output was much larger in 2020 than in 2008 and 2009 during the financial crisis, the IMF
cited several mitigating factors limiting the long-term damage this time around:

-facing sectors. These

are typically at the end of supply chains and tend to be less influential in affecting
economy-wide productivity than upstream businesses.

49
ONS, Homeworking hours, rewards and opportunities in the UK: 2011 to 2020, April 2020.

69 Fiscal risks report

Coronavirus pandemic

Government support has been far greater. Advanced economy governments have spent
an average of 8 per cent of GDP in supporting households, businesses and public
services in 2020,50 compared to an average of just 2 per cent of GDP in 2009.51

Financial instability has been largely avoided. Such instability has historically been
associated with deeper and longer-lasting recessions, but it has been avoided through a
combination of post-2008 reforms to financial regulation alongside the provision of
prompt and extensive support by governments and central banks.

2.56 -
of vaccine rollout explain much of the cross-
scarring, with the average loss of output running from less than 1 per cent in the advanced
economies to 4 per cent in emerging markets and 6 per cent in developing countries. For
the UK, the latest WEO projection includes a scarring effect of 4 per cent 1 percentage
point greater than our own scarring assumption.

2.57 In its May 2021 Economic Outlook, the OECD estimated that potential output in the UK
would be 2 per cent lower in 2022 than its pre-pandemic forecast. The OECD also
calculated that external forecasters had revised down the level of GDP in the UK in 2025 by
an average 3.8 per cent relative to pre-pandemic projections.

2.58 The Bank of England expects scarring of around 1¼ per cent at its three-year forecast
horizon. This is expected to come mainly through the productivity channel, which partly
comes through weaker business investment lowering the capital stock. The Bank also
expects weaker TFP growth as a result of the lower investment and the lack of skills
improvement by those who have not been working during the pandemic. As outlined in our
March 2020 EFO, the Bank had a weaker projection for potential output than we did before
the onset of the pandemic. This lessens the gap between our overall potential output
forecasts relative to the gap between our respective scarring assumptions.

2.59 Of course, what matters for the sustainability of the public finances is the overall outlook for
-term
GDP forecasts will reflect a combination of pandemic effects, Brexit effects, and assumptions
about underlying potential output growth all of which are highly uncertain. Comparing
our GDP forecast from March to the five-year forecasts compiled by the Treasury in May,
our central forecast is towards the bottom of the range in the near term reflecting the
smaller shortfall in output in recent months than we assumed in March (Chart 2.23). In the
medium term, our forecast remains towards the middle of the range of forecasts and, after
five years, is only 0.5 per cent below the overall average and 0.8 per cent below the
average of those new forecasts produced in May. This difference is relatively small given the
uncertainty surrounding economic forecasts in the current environment, as illustrated by our
upside and downside scenarios. In 2025, our upside scenario is close to that of the most
optimistic external forecaster and our downside scenario is slightly below the most
pessimistic one, suggesting they continue to provide a plausible range for future outcomes.

50
Figure 1.7 of IMF, Fiscal Monitor: A Fair Shot, April 2021.
51
Table 3.4 of IMF, Fiscal Implications of the Global Economic and Financial Crisis, June 2009.

Fiscal risks report 70

 Coronavirus pandemic

Chart 2.23: Comparison of GDP forecasts

110

105

100
2019 = 100

95

Range of external forecasts

OBR November 2020 upside scenario
90 OBR November 2020 downside scenario
Bank of England
New independent forecasts
Independent average
OBR March 2021 forecast
85
2019 2020 2021 2022 2023 2024 2025
Note: Bank of England forecast excludes the backcast.
Source: Bank of England, HM Treasury, ONS, OBR

Future developments
2.60 The eventual extent of scarring is still highly dependent on the path of the pandemic in the
coming months and on policy responses to it. We outlined some of the risks to the
epidemiological assumptions that underpinned our latest forecast in Box 2.1 of our March
2021 EFO, most of which are still relevant. Indeed, the delta variant has led to the
Government announcing a four-week delay to 19 July
for lifting the remaining public health restrictions in England, which underpinned our March
forecast. Another major scarring-related uncertainty is how businesses and households
respond to the withdrawal of government support measures, much of which currently
remains in place.

2.61 As is our usual practice, we will review our potential output assumptions, including
pandemic-related scarring, and revise them, if appropriate, in our next EFO and in
subsequent forecasts as more information accrues. As time goes by, however, it will become
increasingly difficult to distinguish the effects of the pandemic on the economy from those
caused by other factors such as Brexit or the general stagnation in productivity since the
financial crisis. Nonetheless, some of the information we will be reviewing in coming
months to inform our potential output forecasts are set out in Table 2.2. It includes:

Data on the performance of the labour market after the CJRS closes at the end of
September. The latest data show that in May 2021 there were still 2.6 million people
on the CJRS. While this is significantly down from the peak of 8.7 million in April
2020, it is still around 5 per cent of the adult population (see Box 2.1). While
unemployment and inactivity have so far not risen significantly, the extent to which
those on furlough flow into each could materially affect the extent of labour market

71 Fiscal risks report

Coronavirus pandemic

scarring. However, the full extent of labour market scarring will depend on the ability
of the jobless to subsequently move into new jobs, sectors and occupations.

The extent and composition of firm insolvencies. Since the beginning of the pandemic,
insolvencies have been remarkably subdued. Some of this will be a product of the
grants and business rates holidays,
guaranteed loans, and in large part paying the wages of furloughed workers. The
Government also introduced a temporary directive in April 2020 restricting the use of
winding-up petitions, which has been subsequently extended until the end of
September 2021, reducing the possibility of insolvencies until then. Eviction protection
for commercial tenants has since been extended to March 2022. These factors help to
keep some otherwise viable firms from failing, supporting productivity by maintaining
firm-specific capital and knowledge. However, these protections may also have had an
offsetting adverse impact on productivity by keeping otherwise unviable businesses
operating (so-
additional protections end, both the extent and composition of firm insolvencies may
provide some indication of the scarring of productivity.

The recovery in business investment and any revisions to historical data. Business
investment data are always prone to revision and the ONS has emphasised the
increased uncertainty around data caused as a result of the pandemic. The latest
vintage of data show that business investment in the first quarter of 2021 was still 17
per cent below the pre-pandemic peak, lagging the recovery in GDP, which was only 9
per cent below. The outlook for the continued recovery is further clouded by the
uncertainty around the effect of the temporary super-deduction capital allowance and
increase in the corporation tax rate that were announced in the Budget in March. The
former is likely to have a significant effect on the timing of investment, although the
size is particularly uncertain given its lack of precedent. Evidence from the Bank of

10 per cent increase that we incorporated in our forecast.52 Business investment data
will give an indication of the productivity scarring effect through capital shallowing, but
its volatility and tendency for to be revised significantly between releases will inevitably
cloud the picture.

Data on net migration during the pandemic and indications about the extent to which
those will return to the UK. In July, the ONS will be reweighting its
labour market statistics using RTI data to give a timelier view of the UK population.
They will also be updating their modelling of net migration estimates later this year
and providing 2020-based mid-year population projections. These will give additional
information on migration, but it is unlikely that we will have a robust estimate of the
UK population until the latest Census results are released in 2022. Even then, there will
still be considerable uncertainty on the prospects for net migration in the medium term,
including how many of those who left will return to the UK. This is compounded by the
fact that any catch-up immigration will need to take place under the new post-Brexit

52
Bunn et al, Influences on investment by UK businesses: evidence form the Decision Maker Panel, 25 June 2021.

Fiscal risks report 72

 Coronavirus pandemic

immigration system, which is tighter than its predecessor for those entering the UK
from EU member states.

Table 2.2: Forthcoming scarring-related data releases

Release Information Date
May Labour Force Survey (LFS) Population scarring based on RTI data 15 July 2021
1st estimate of Q2 2021 GDP Business investment recovery (capital shallowing) 12 August 2021
Q2 2021 Quarterly National Accounts,
Pace of recovery and capital shallowing 30 September 2021
incorporating 2021 Blue Book revisions
September real time information (RTI) Labour market during last month of CJRS 12 October 2021
August 2021 GDP Output post lifting of final covid restrictions 13 October 2021
September LFS Labour market during last month of CJRS 16 November 2021
October RTI 1st data on labour market post-CJRS 16 November 2021
Capital scrapping and TFP effects as
Monthly Insolvency Statistics 16 November 2021
government support fades
October LFS Detailed data on labour market post-CJRS 14 December 2021
ONS migration modelling and 2020-
based population projections Indication of population scarring Late-2021

73 Fiscal risks report

Coronavirus pandemic

Box 2.1: The CJRS and unemployment

In the March 2021 Budget, the Chancellor announced that the CJRS will be fully withdrawn at
the end of September, and that employer contributions will increase in monthly steps from July
onwards. Our forecasts for GDP, unemployment, and other key fiscal determinants assume that
the vast majority of those still on furlough will be able to return to their jobs or find alternative
employment. The latest data show that there were still 2.6 million people on the CJRS in May
2021, with the number falling quickly in recent months as the economy has reopened and
activity rebounded. Even so, jobs fully or partly furloughed still account for shortfall in labour
utilisation relative to the month prior to the pandemic.
Chart A: Change in employment-related indicators during the pandemic
0

-1
Cumulative change in employment (per cent)

-2

-3

-4

-5

-6
Payrolled employee jobs
-7
Fully furloughed
-8
Partially furloughed
-9
Self-employed
-10
Feb 20 Apr 20 Jun 20 Aug 20 Oct 20 Dec 20 Feb 21 Apr 21

Note: May 2021 self-employment data (shaded) is estimated based on BICS data, as LFS is not yet available.
Source: HMRC, ONS, Resolution Foundation, OBR

While the numbers furloughed are down by 5.1 million from the peak of 8.7 million at the
height of the first lockdown in April 2020, it is still a very large programme, paying a large
proportion of the wages of 9 per cent of all payrolled employees in the UK. The latest ONS BICS
data suggest that this proportion has fallen a little further to 7 per cent by mid-June.
Given the large number of people still on the scheme, and their growing concentration in a few
of the hardest hit sectors, there remains considerable uncertainty as to how many will be able to
return to their previous roles or employers, and how many will need to look for other
employment. Chart B shows the proportion of furloughed employees and the vacancy rate in
each sector as of May 2021. It shows that furloughed employees are increasingly concentrated
in a few sectors, with accommodation and food services and arts and entertainment being the
most affected. Together, these two sectors accounted for over a third of the 2.6 million
furloughed employments in May 2021, up from around a quarter during the first lockdown.
The capacity for these sectors to fully reabsorb furloughed workers over the next few months as
the scheme is wound down will depend, in part, on how quickly the remaining public health

Fiscal risks report 74

 Coronavirus pandemic

restrictions affecting these sectors and international travel can be lifted. It will also depend on
how sustainable the rebound in social consumption seen in the wake of the third lockdown
proves. The rush to fill positions as these sectors reopened has led to them registering the highest
vacancy rates at over 6 per cent, higher even than before the pandemic. This is encouraging, if
tentative, support for our assumption that most furloughed employees will find work quite
quickly, though considerable uncertainty remains.
As of May 2021, over 30 per cent of employees in both sectors remained on furlough, so a
significant part of the reabsorption process has yet to occur. And economy-wide, the vacancy
rate remains below the 3 per cent pre-pandemic average, while around 9 per cent of all
payrolled employees were still furloughed in May 2021. All these indicators point to this still
being a relatively early stage of the post-pandemic adjustment in the labour market.
Chart B: Sectoral breakdown of furloughed employees and vacancies (May 2021)
7
Accommodation
Vacancies as a share of total employments (per cent)

Arts and and food services

6
entertainment

4
Whole economy
Other services
3

2
Wholesale and retail
trade
1

0
0 5 10 15 20 25 30 35 40 45 50
Employments furloughed as a share of total employments (per cent)
Note: The size of the bubbles represent the sectoral share of total employees furloughed.
Source: ONS, HMRC, OBR

Longer-term fiscal legacy of the pandemic

2.62 The most important long-term fiscal impacts of the pandemic are likely to be those that flow
from any scarring of potential output. But there are other factors that could have long-term
fiscal implications. Here we discuss two: demographic developments; and receipts-specific
scarring.

Longer-term fiscal risks from demographic developments

2.63 Excess deaths due to coronavirus have a direct effect on public spending. The state pensions
caseload is estimated to be 0.8 per cent lower as a result, which reduced pensioner
spending by £0.9 billion in 2021-22 in our March forecast. Our medium-term forecast

75 Fiscal risks report

Coronavirus pandemic

assumes that these excess deaths reflect lives being cut short, rather than a permanent
change in mortality rates at older ages, with correspondingly fewer deaths occurring in
subsequent years until mortality rates get back to pre-pandemic assumptions. This process is
assumed to take eight years, so extends beyond our medium-term forecast horizon.53

2.64 The pandemic and Brexit may also have implications for the size and age profile of the UK
population, which could have long-run fiscal implications. The age profile of the 128,000
coronavirus deaths recorded to date has been heavily concentrated among people of ages
that are associated with net fiscal costs (Chart 2.24). That reflects them paying less tax and
no National Insurance, while receiving more in health and social care services, and state
pensions and other social security benefits. Indeed, the ONS estimates that around 42,000
coronavirus deaths were care home residents.54

Chart 2.24: Coronavirus deaths versus net fiscal costs by age group
40

30

20

10
Thousand

-10

-20
Coronavirus deaths
-30
Net annual fiscal cost (£)
-40
<1 1 to 5 to 10 to 15 to 20 to 25 to 30 to 35 to 40 to 45 to 50 to 55 to 60 to 65 to 70 to 75 to 80 to 85 to 90
4 9 14 19 24 29 34 39 44 49 54 59 64 69 74 79 84 89 plus
Age group
Source: HMRC, OBR

2.65
change:

The previous section discussed what we know so far about the effect of the pandemic
on net migration and what that means for potential output. Changes in net migration
also have longer-term fiscal consequences due to the different age profile of net
migrants relative to the native population, which means that net inward migration
typically lowers the old-age dependency ratio and improves fiscal sustainability. In our
2018 Fiscal sustainability report (FSR), raising or lowering annual net inward migration
by 80,000 a year (in line with the high and low migration variants of the ONS

53
Living with Covid-19: balancing costs against benefits in the face of the virus, Miles, Stedman and Heald, July 2020.
54
Care home resident deaths registered in England and Wales, provisional, ONS, June 2021.

Fiscal risks report 76

 Coronavirus pandemic

population projections available at the time) led to debt being 11 per cent lower after
50 years in the high variant and 14 per cent higher in the low variant.

The number of births fell sharply in December and January, nine months after first
lockdown, but picked up again in February and March of this year, nine months after
that lockdown was eased. It is therefore not clear at this stage whether the pandemic
will have lasting effects on the existing downward trend in fertility rates. The fiscal
consequences of changes in birth rates change as the affected cohorts age fewer
births would initially lower spending while they are children, then lower receipts (and
output) when they are in work, then finally lower spending again when they retire.

Potentially the greatest long-term uncertainty relates to any legacy impact on mortality
rates. As noted, our forecast assumes that excess deaths to date were all brought
forward from future years, with no lasting effects on the pre-pandemic trend for
mortality rates to continue declining steadily over time. Ageing is a key long-term
pressure on the public finances, so if coronavirus continues to circulate and, despite
high vaccine take-up and efficacy, leads to higher than previously assumed mortality
at older ages, that pressure would be reduced though only modestly unless the
lasting effects on mortality rates were very severe. In our 2018 FSR, we tested the
sensitivity of spending on state pensions and other pensioner benefits to different
assumptions about life expectancy. Varying it by roughly 10 years either side of the
baseline assumption (in line with the ONS old-age and young-age structure
population variants at the time) left spending on these items in 2067-68 up or down
by around ¾ per cent of GDP relative to baseline spending of 8.2 per cent of GDP.

2.66 Similarly to the challenge of estimating scarring of potential output while fiscal support
measures remain in place, it will be difficult to determine whether any of the demographic
changes witnessed during the pandemic will have lasting effects. This makes the next (and
probably subsequent) ONS population projections much more important and uncertain than
usual, with the potential to affect our assessment of both the medium- and long-term fiscal
outlook materially when they are published. They have been delayed so that they can reflect
the Census. This will add to forecast uncertainty in the intervening period and will mean
there is a risk of significant revisions when they can be incorporated in our forecasts.

Longer-term receipts-to-GDP ratio scarring

2.67 In addition to the permanent effect on receipts from long-term economic scarring, some
long-term pressures on revenue from tax bases have been accelerated by structural shifts
brought about by the pandemic. These include a move to online retail reducing demand for
retail premises, and the shift towards working from home reducing
demand for office-based space. Both these trends could hit future business rates receipts,
which our latest forecast predicts will raise £35.0 billion in 2025-26. Even a relatively
modest 5 per cent shortfall would therefore take £1.8 billion off receipts in that year.

2.68 The accelerated digitalisation of economic activity also poses difficulties in what can be
taxed and where a subject we discussed in Chapter 4 of our 2019 FRR and where

77 Fiscal risks report

Coronavirus pandemic

lockdowns prompted several years worth of the pre-pandemic trend towards online retail to
take place in a single year. The share of retail sales taking place online jumped by 16
percentage points between February 2020 and a peak of 36 per cent in February 2021,
before declining to 28 per cent in May as retail settings reopened. But that 8 percentage
point rise relative to the pre-pandemic position is still six times greater than the average
annual rise of only 1.3 percentage points recorded over the preceding decade. This shift
online could be associated with a rise in the tax gap for VAT. A 5 per cent rise in the VAT
gap (equivalent to 0.4 percentage points in 2025-26 in our latest forecast) would lower
receipts by £0.6 billion in 2025-26.

2.69 Other tax bases at risk include:

Fuel duty. Less use of public transport could boost fuel duty if it results in an increase in
driving to work, whereas more working from home could have the opposite effect. If
receipts were to settle 5 per cent lower than our forecast assumes, the shortfall would
be £1.6 billion in 2025-26. That said, pandemic-related risks are modest relative to
trends towards electric vehicles as a result of regulations to help deliver a net-zero
economy by 2050 (see Chapter 3). They are also small relative to the policy risk
signalled by fuel duty rates having been frozen at every Budget of the past decade.

VAT. Receipts could be permanently reduced relative to GDP if, for example, people
do not return to eating in restaurants, cafes, and pubs to the same extent as pre-
pandemic, instead consuming shop-bought food that is more likely to be zero-rated
rather than standard-rated for VAT. If 5 per cent of VAT decl
permanently lost to zero-rated spending, VAT receipts in
2025-26 would be around £0.4 billion lower.

Air passenger duty. Our forecast assumes that restrictions on overseas travel, and
reduced consumer and business preferences for air travel, will result in air passenger
duty receipts remaining around 10 per cent below the pre-pandemic path in five years
time. Scarring effects are particularly uncertain here, though the tax itself is a small
revenue source. Even a further shortfall of 10 per cent relative to our central forecast
would only take £0.4 billion off receipts in 2025-26.

2.70 Table 2.3 summarises these potential sources of revenue scarring and how they would
affect the receipts-to-GDP ratio were they all to crystallise in this way.

Table 2.3: Potential long-term impacts on the receipts-to-GDP ratio

Effect in 2025-26
Revenue risk £ billion Per cent of GDP
Business Rates - 5 per cent shortfall 1.8 0.07
VAT gap - 5 per cent rise 0.6 0.02
Fuel Duty - 5 per cent fall 1.5 0.06
VAT in Food and beverages service sector - 5 per cent fall 0.4 0.02
Air passenger duty - 10 per cent fall 0.4 0.02
Total 4.7 0.18

Fiscal risks report 78

 Coronavirus pandemic

Lessons from coronavirus for other large fiscal risks

2.71 As the largest fiscal risk to have crystallised in the post-war era, the coronavirus pandemic
also provides valuable insights as to the nature of catastrophic risks and how economic
forecasters and policymakers can better understand and respond to them. The experience of
the past eighteen months points us to the following lessons for the handling of other large
fiscal risks, including those discussed elsewhere in this Fiscal risks report.

1 Catastrophic risks are real and may have become more frequent. Just two decades
into this century, the UK and other advanced economies have now experienced two
growing financial leverage,
economic interdependence, and other manmade risk factors may make future shocks
both more frequent and more severe. Producers and users of economic and fiscal
forecasts tend to focus on a central view of medium-term prospects in which output
returns to a judgementally determined trend as the effects of past shocks dissipate. But
it is equally arguably, more important to focus on the risks around that forecast
that arise from inevitable future shocks. Forecasters should do more to emphasise the
uncertainty surrounding both near- and longer-term economic and fiscal prospects.

2 Economic shocks affect both supply and demand. Macroeconomic forecasting and
analysis rely on being able to evaluate the effect of a shock or indeed any news on
both supply and demand and whether those effects are likely to be persistent or
transitory. While conventional cyclical shocks affect mainly demand, recent shocks
the financial crisis, Brexit and the pandemic have materially affected both supply and
demand. This has exposed how poorly supply-side developments are understood,
measured and modelled relative to textbook business cycle fluctuations in demand.
Forecasters need to raise their capacity to assess and monitor both the immediate and
longer-term supply-side impact of novel shocks and any policy response.

3 Global interconnectedness can be both an asset and a liability. As one of the most
globally connected economies, the UK is highly exposed to risks emanating from
abroad in the form of not only pandemic disease but also other forms of economic

investment also made it a world leader in development, production, and rollout of one

-
based economy to continue to operate through the pandemic and the Government to
deliver timely fiscal support, but also renders the economy vulnerable to cyberattacks
on critical IT infrastructure the potential fiscal risks of which are discussed in Box 5.1.

4 While it may be difficult to predict when catastrophic risks will materialise, it is possible
to anticipate their broad effects if they do. The risk of a global pandemic was on the
top of government risk registers for a decade before coronavirus arrived but attracted
relatively little (and in hindsight far too little) attention from the economic community.
However, both the experience from previous epidemics such as the 1918 flu, Ebola,
and SARS, and modelling by the US Congressional Budget Office and the World Bank,

79 Fiscal risks report

Coronavirus pandemic

provided clear indications of where and how badly economies might be affected, even
though both modelled an influenza rather than coronavirus pandemic. In 2008 the
World Bank estimated that a severe and moderate flu pandemic could reduce global
GDP by 4.8 per cent and 2 per cent respectively,55 compared to the 3.3 per cent fall in
2020 due to the coronavirus pandemic. The CBO estimated US GDP losses of 4¼ per
cent in a severe flu pandemic compared to the 3.5 per cent fall recorded last year.56

5 When investing in risk prevention, governments tend to only . In the

decade following the 2008 financial crisis, significant resources were dedicated to
improving the oversight and resilience of the financial sector, which paid dividends
during the pandemic by helping to prevent it from triggering another financial crisis.
And East Asian countries that invested in epidemic surveillance following the SARS and
MERS outbreaks were more capable of combating the pandemic from the beginning.
However, the 2016 report of the UN High-level Panel on Global Response to Health

57
woefully insufficient
The difficulty in anticipating the precise timing
and nature o -
scanning and investing in generic risk management systems and structures.

6 There are significant advantages in preventing or halting a process that involves

rapidly escalating costs early. While economic theory and practice emphasises the
option value of delaying decisions, this can be suboptimal in the face of rapidly
escalating costs. Countries that acted quickly to contain the spread of the virus
experienced fewer deaths, shallower recessions, and earlier economic recoveries.
These countries did not necessarily see lower fiscal costs from the pandemic, but more
of their increase in borrowing was due to discretionary fiscal policy rather than as a
result of the decline in output or pressures on their health systems and is therefore
more likely to prove reversable.

7 People appear willing to make sacrifices for a clearly-defined public good. In the early
stages of the pandemic, there was concern about defiance or fatigue in relation to
public health restrictions and requirements. In fact, compliance with public health
restrictions remained high throughout the pandemic in the UK and vaccine take-up
also exceeded expectations. In total, the UK experienced a 10 per cent loss of output
and committed 12 per cent of GDP in public funds in order to combat the pandemic in
2020. The annual economic and fiscal costs of tackling other potential catastrophic
risks, like climate change, are likely to be just a fraction of this.

8 Economies can sometimes adapt remarkably quickly to structural changes. While the
initial shock associated with the pandemic and initial lockdowns was greater than
many economists predicted, they were also surprised by the speed and strength of the
subsequent recovery in economic activity (including its resilience during subsequent

55
Evaluating the Economic Consequences of Avian Influenza, World Bank, 2008
56
A Potential Influenza Pandemic: Possible Macroeconomic Effects and Policy Issues, Congressional Budget Office, 2006
57
Protecting humanity from future health crises: Report of the high-level panel on the global response to health crises, United Nations,
2016.

Fiscal risks report 80

 Coronavirus pandemic

lockdowns). The contribution of prior investments in information technology that

enabled people to work, shop, learn, and be entertained online was critical to
enabling this transition, as was fiscal policy in allowing households and firms to
maintain consumption, employment, and liquidity through the transition.

9 Fiscal policy can and needs to be more nimble than was previously thought. Before the
pandemic, one of the central preoccupations among macroeconomists was that
monetary policy had been exhausted as the principal instrument for managing
fluctuations in aggregate demand but fiscal policy could not act with the speed and
scale necessary to prevent lasting damage to the economy. In fact, across advanced
economies the pandemic induced a fiscal policy response unprecedented in its speed,
scale, and novelty. While this added 18.7 per cent of GDP to the debts of the average
advanced economy by the end of 2021, it also prevented the much greater economic
costs associated with the deeper, longer, and more disruptive economic contraction
that could have resulted from not intervening.

10 In the absence of perfect foresight, fiscal space may be the single most valuable risk
management tool. Throughout its history, the UK government has relied on its ability to
borrow large sums quickly in order to respond comprehensively to major economic
and security threats. It was able to do so courtesy of its relatively low levels of public
indebtedness, deep and liquid domestic capital market (supported by monetary policy),
and by benefitting whenever there has been a general flight to safety. Fiscal
policymakers must trade-off making significant investments in the prevention of
specific potential threats with preserving sufficient fiscal room for manoeuvre to
respond to those risks which it did not anticipate or could not prevent.

81 Fiscal risks report

Fiscal risks report 82
3 Climate change

Introduction
3.1 Climate change threatens lives and livelihoods around the world. While its effects are
unevenly distributed, even countries such as the UK that will be relatively less affected in the
first half of this century would still suffer greatly if unmitigated global warming continued
indefinitely.1 the defining crisis of our
time No corner of the globe is immune from [its] devastating

2
water insecurity, economic disruption, conflict, and terrorism.

3.2 Governments and public alike have acknowledged these threats. Targets for limiting global
warming were agreed in Kyoto in 1997, in Copenhagen in 2009, and most recently and
comprehensively in Paris in 2015. The UK Government has since legislated to achieve net
zero emissions by 2050 one of 131 countries that have either made net zero commitments
or have targets under discussion, but one of only six to have so far put that commitment into
legislation.3 This has not yet translated into falling global emissions (with the exception of
years of economic crisis), although the rate of emissions growth has slowed somewhat since
the 1980s relative to the rapid increases recorded in the post-war decades. Emissions in
major advanced economies are either broadly flat or have started to fall, but emissions
from emerging markets continue to rise, in part reflecting the relocation of industrial
production from the advanced economies who continue to consume its outputs.4

3.3 The fiscal implications of climate change for the UK are complicated and depend upon the
policy response at home and abroad. Global trends that are largely beyond the UK

with adapting to the changes that brings. Unmitigated climate change would ultimately have
catastrophic economic and fiscal consequences, but even meeting Paris goals implies some

emissions reductions, the fiscal risks from being left behind in the global decarbonisation
process have als
costs of mitigating emissions in the transition to net zero by 2050 and the extent to which
opportunities from associated technological advances can be grasped. There are many

1
For instance, Swiss Re estimates that out of 48 major economies, the UK will be the 15th least affected by climate change by the middle
of the century, Swiss Re Institute, The economics of climate change: no action not an option, April 2021. And a World Bank study reached
the same conclusion, scoring the UK in the bucket of countries most resilient to, and most insulated from, the transition to a low-carbon
economy due to relatively low levels of dependence on the domestic consumption and export of fossil fuels, World Bank, Diversification
and cooperation in a decarbonizing world: climate strategies for fossil-fuel dependent countries, World Bank, 2020.
2
United Nations, The Climate Crisis A Race We Can Win, 2020.
3
Energy & Climate Intelligence Unit, Net Zero Emissions Race: 2021 scorecard, 2021.
4
European Commission, Fossil CO2 emissions of all world countries, 2020.

83 Fiscal risks report

Climate change

possible paths ahead, each with different fiscal implications. Uncertainty around them all is
pervasive and will depend on choices made at each stage.

3.4 With that in mind, this chapter:

discusses the science of climate change, its potential economic impact, and the key
policy levers available to mitigate it;

illustrates the potential physical, economic, and fiscal risks to the UK from different
paths for global warming;

outlines costs of decarbonising the UK economy by 2050, drawing on projections

presented by the Climate Change Committee;

considers the potential implications of making the transition to net zero emissions for
public spending and revenues;

presents a set of fiscal scenarios for achieving net zero emissions under different
assumptions; and

draws conclusions.

3.5 It is important to stress that the quantification of fiscal risks in this chapter is largely
illustrative. This represents the first step in a programme of work to refine our understanding
of how climate-related fiscal risks propagate and the size of their potential effects.5 One key

emissions target, which it plans to describe more fully in a Net Zero Strategy later this year.6
Where long-term policies have yet to be set and consistent with the Charter for Budget
Responsibility we have made assumptions about the tax and spending implications of
different climate paths and different scenarios for bringing emissions to net zero.7

Climate change science, economics, and policy

3.6 In our 2019 Fiscal risks report (FRR), we developed a framework for assessing the fiscal risks
to the UK from climate change t
climate-related risks to financial stability. It split risks into those stemming from climate

economy, including the policies necessary to achieve

5
The Congressional Budget Office our equivalent in the US has launched a similar programme. In September 2020 it published
, a sophisticated, if partial, assessment of how different climate
paths in the US might affect real GDP in the period to 2050. In April 2021 it published Budgetary Effects of Climate Change and of
Potential Legislative Responses to It, an initial and largely descriptive discussion of how the US federal budget might be affected.
6
See Department for Business, Energy and Industrial Strategy, Impact Assessment for the sixth carbon budget, April 2021, which states
The government will publish the Net Zero Strategy later this year, setting out its vision for transitioning to a net zero economy. This will
economy. These
sectoral plans include the Energy White Paper published last December, the Industrial Decarbonisation Strategy published in March, as well
as the Transport Decarbonisation Plan, Hydrogen Strategy and Heat and Buildings Strategy to be published shortly.
7
Paragraph 4.15 of the Charter that was passed by Parliament in January 2 where a long-term policy has not yet
been set by the government, the OBR will set out the assumptions it makes in its projections regarding policy transparently

Fiscal risks report 84

 Climate change

framework, it is helpful first to consider some fundamental features of the science of global
warming (which underlies the physical risks), the microeconomic determinants of emissions
(which influence the transition risks), and their combined macroeconomic impact (which
reflects both risks in different combinations depending on the path taken).

Greenhouse gas emissions and global warming

3.7 te Change (IPCC) estimates that human activity
has led to a rise in average temperature of roughly 1°C relative to pre-industrial levels
8
due to the effects of greenhouse gases emitted into the
9
atmosphere. But this past activity is unlikely to generate significant further warming in the
future as it is the stock of greenhouse gases in the atmosphere, rather than the annual flow
of new emissions, that determines the average temperature. Since the atmospheric lifetime
of most greenhouse gases is very long, it follows that the net flow of new emissions must be
eliminated if the temperature is to be stabilised. So it is future human activity and emissions
that will largely determine how much further temperatures rise. That is why the Paris goal of
-industrial levels requires emissions to be
thereby stabilising the stock of emissions and global temperatures.10

3.8 Chart 3.1 plots the levels of carbon dioxide (CO2, the most important greenhouse gas in
terms of its aggregate effect on global temperatures) emitted in three of the scenarios
produced by the Network for Greening the Financial System (NGFS),11 plus a benchmark
IPCC scenario in which global warming is completed unmitigated. It shows that even though
currently implemented policies may be enough to stop the flow of emissions increasing, they
are not enough to stop the stock of CO2 in the atmosphere increasing. The stock only begins
to be reduced in the two scenarios containing sharp cuts to net emissions it is only these
emissions trajectories that would imply a high probability of limiting warming to the Paris
target of -industrial levels.

8
IPCC, Global Warming of 1.5°C. An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and
related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change,
sustainable development, and efforts to eradicate poverty, 2018.
9
The main gases responsible for the greenhouse effect include carbon dioxide, methane, nitrous oxide, and water vapor (which all occur
naturally), and fluorinated gases ( - which are synthetic). Their effects on global warming depend on their concentration in the
atmosphere, how long they remain in the atmosphere after being emitted, and their effectiveness at trapping heat in the atmosphere.
10
See Article 4 of United Nations, Paris Agreement, 2015.
11
The NGFS is a network of central banks and supervisors of financial institutions launched in 2017 to contribute to the development of
climate risk management in the financial sector. One of its workstreams has been overcome what was seen as a major obstacle to
undertaking climate risk analysis: the lack of detailed scenarios that consider both the physical and transition risks from climate change,
and their economic impacts. The challenges and costs of creating such scenarios were felt to be beyond most individual firms or
institutions (as they would be for most fiscal watchdogs too). The NGFS has developed a common set of scenarios to fill that gap.

85 Fiscal risks report

Climate change

Chart 3.1: Carbon dioxide emissions and atmospheric concentration

180 1000
Global CO2 emissions Atmospheric concentration of CO2
160
140
800
Current policies

CO 2 parts per million

120
Gigatonnes of CO 2

Delayed transition
100
Net zero 2050
80 600
Unmitigated warming
60 Outturn
40
400
20
0
-20 200
1950 1970 1990 2010 2030 2050 2070 2090 1950 1970 1990 2010 2030 2050 2070 2090
Source: Our World in Data, NGFS Climate Scenarios Database, GCAM model, International Institute for Applied Systems
Analysis RCP database, Joint Global Change Research Institute GCAM database, and OBR calculations.

3.9 Predicting future warming using the change in the stock of greenhouse gases in the
atmosphere is not straightforward. In particular, the further temperatures rise, the greater
the increased risk of triggering tipping points at which adverse feedback loops and cascade
effects kick in. This could involve, for example, accelerated melting of the Greenland ice cap
causing temperatures to rise faster as it reflects less heat back out of the atmosphere,
thereby triggering faster degradation of the Siberian permafrost, in turn releasing more
greenhouse gases into the atmosphere and causing further temperature rises.12 The
associated uncertainty is illustrated in Chart 3.2, which plots median temperature rises and
the risk of temperature increases of 6°C or more against a measure of the stock of
greenhouse gas (in this case, carbon dioxide).13 In the extreme climate scenarios that these
tipping points and cascade effects could generate, it would be nature, rather than human
action, that ultimately brings net emissions towards zero by leading to depopulation.14

12
Steffen, W., Rockström, J., Richardson, K., Lenton, T., Folke, C., Liverman, D., Summerhayes, C., Barnosky, A., Cornell, S., Crucifix, M.,
Donges, J., Fetzer, I., Lade, S., Scheffer, M., Winkelmann, R. and Schellnhuber, H., Trajectories of the Earth System in the Anthropocene,
July 2018.
13
As reported in Wagner, G., and Weitzman, M., Climate shock: the economic consequences of a hotter planet, 2015, Table 3.1. The
median temperature increases are based on an assumed climate sensitivity that global temperatures increase by 2.6°C every time the
atmospheric concentration of greenhouse gases doubles. The probabilities that temperature rises exceed 6°C are calculated by taking the
2 to 4.5°C likely Fifth Assessment Report, assuming that this
parameter has a log- likely
likely
- ue is 2.6°C.
14
Ekins, P. and Zenghelis, D. The costs and benefits of environmental sustainability, Sustainability Science, March 2021.

Fiscal risks report 86

 Climate change

Chart 3.2: Median temperatures versus the risk of exceeding 6°C temperature rises
at different atmospheric concentrations of greenhouse gases
18 18
Median temperature increase (left axis)
16 16
Chance of >6°C temperature increase (right axis)
14 14
Temperature increase (°C)

12 12

Probability (per cent)

10 10

8 8

6 6

4 4

2 2

0 0
400 450 500 550 600 650 700 750 800
CO 2 equivalents concentration (parts per million)
Source: Wagner, G., and Weitzman, M., Climate shock: the economic consequences of a hotter planet, 2015, Table 3.1

Market failures and greenhouse gas emissions

15
3.10 As the
global cost of emissions to current and future generations is not borne by the producers of
those emissions, too much is generated from a social perspective. If that were the only
market failure present, then a solution would be to impose an appropriate carbon tax to
internalise the costs for producers, with the main challenge for policymakers being choosing
the appropriate tax rate, imposing it domestically, and implementing it consistently across
the world. In practice, the presence of other market failures and distortions, together with
uncertainty about the effectiveness of different policy interventions and their impact across
society, means that a mix of policies is warranted.

3.11 These other market failures and distortions include:16

Positive externalities. The developers of new technologies are rarely able to capture all
the gains, potentially leading to underinvestment in what is essentially a quasi-public
good. That is particularly likely to be the case when the technologies in question have
long horizons. Indeed for removals technologies, the benefits of investment may be so
hard to capture that development would be minimal without government involvement.

Incomplete markets
but would bear the costs of future global warming. Consequently, their preferences

15
Pigou, A.C., The Economics of Welfare, 1920. In addition to their effects on global warming, emissions may also lead to other negative
externalities, such as the impacts on health outcomes from air pollution caused by burning fossil fuels.
16
See, for example, Annex B of HM Treasury, Net Zero Review: Interim report, December 2020.

87 Fiscal risks report

Climate change

Information failures. The costs and benefits of some technologies are poorly
l bills from energy efficiency
measures, potentially leading to underinvestment (as discussed later in Box 3.3).

Credit constraints. Frictions in financial markets that limit access to finance to fund
investment by prospective producers in unproven green technologies or by poorer
households in the deployment of proven ones.

Network effects and coordination failures. Many green technologies will be cheaper
and more effective if widely adopted, but achieving widespread adoption requires
breaking out of the existing-technology equilibrium. Since no individual actor has a
sufficiently strong incentive to achieve that, it may be desirable for governments to
steer businesses and households towards particular models and standards. Some
commentators have therefore emphasised how the structure of the industries involved
in the transition to net zero might lead to path dependencies and multiple equilibria,
enabling the state to help society coordinate on better outcomes than market actors
would reach by themselves.17 For instance, provision of charging infrastructure to

expectations. Or temporary investments to overcome initial costs and inertia in the

power sector, enabling a permanent shift in economies to greener networks.

The macroeconomic impact of climate change

3.12 Beyond a certain point, hotter temperatures potentially seriously damage the economy. And
warming in general can affect the economy through numerous channels, including by
lowering crop yields and impairing the productivity of the workforce. A study published in
the scientific journal Nature estimates that productivity peaks at an average annual
temperature of around 13°C, and that in the extreme, unmitigated emissions scenario
shown above (where average temperatures rise by 4°C by 2100) the average level of per
capita global GDP in 2100 would fall by 23 per cent (as the average annual temperature in
most regions is already in excess of the optimal 13°C).18 While the direction of this effect is
relatively clear, the size is uncertain an IMF study looking at an identical climate scenario

3.13 As we highlight in this report, policymakers also need to take the consequences of the
crystallisation of catastrophic risks into account, rather than only focusing on most likely
outcomes. The effects of climate change in other countries are certainly large enough to
trigger catastrophic risks elsewhere, which small open economies like the UK would not be
insulated from for instance, if hotter temperatures in already-hot countries were to lead to
conflict over increasingly scarce water resources, triggering mass migration to more
temperate countries and affecting global supply chains. And climate change is likely to
increase the frequency of extreme weather events, rather than simply raising average
temperatures. We discuss these channels further in paragraphs 3.24 to 3.27.

17
For instance, Ekins, P. and Zenghelis, D. The costs and benefits of environmental sustainability, Sustainability Science, March 2021.
18
Burke, M., Solomon, H.,and Miguel, E., Global non-linear effect of temperature on economic production, 2015.

Fiscal risks report 88

 Climate change

3.14 In addition, mitigating climate change the policy measures taken to reduce net emissions
will also affect GDP. Taxing or banning polluting activities raises the implicit price of

Importantly, the definition of GDP does not incorporate any future benefits from
reducing global warming. So, all else equal, a higher shadow price of carbon reduces GDP
as businesses are encouraged to move away from (privately) efficient carbon-intensive
methods of production towards (socially preferable) lower-carbon methods. This is the case
in both the Bank of England scenarios discussed from paragraph 3.99 onwards.

3.15 While the behavioural response of any particular individual or business to a higher carbon
price may be relatively straightforward to assess, it is difficult to evaluate with any
confidence the impact on the economy as a whole of a structural change as large as the
transition to net zero. There are several channels through which a successful transition could
actually enhance productivity, possibly by more than enough to offset the direct adverse
impact on carbon-intensive activities. For instance, stimulating large-scale investments in
green technologies may have dynamic effects, boosting productivity for all as technology
costs fall.19 In addition, establishing an early dominant position in new green technologies
could create a source of comparative advantage internationally, benefitting future exports.
For this reason, we also include a scenario later in this chapter in which the transition to net
zero raises GDP modestly by 2050.

3.16 We focus in this report on the level of real GDP, rather than the full range of economic
variables affected by climate change. This is to highlight two of the most important channels
through which climate change affects the public finances: the direct fiscal costs of the
transition; and its indirect fiscal impacts via the size of the economy. But of course this
provides only an incomplete picture of the impact of climate change on individuals. First,
the composition of GDP matters, not just its level: the investment required to transition to net

consumption. Second, GDP is measured gross of depreciation, so captures the benefit of

higher investment spending, but not the cost of any resources wasted by prematurely
scrapping parts of the capital stock. Third, climate change and the transition to net zero
may affect prices and interest rates, as well as quantities such as GDP. And finally, and
perhaps most importantly, economic statistics largely capture the economic costs of climate
change only as they crystallise, rather than recognising them upfront. And they do not fully
capture the amenity and common pool value of the climate and other natural resources. But
work to improve the way natural assets are recognised in accounting frameworks is
underway, as Box 3.1 discusses.

19
Again, see Ekins, P. and Zenghelis, D. The costs and benefits of environmental sustainability, Sustainability Science, March 2021.

89 Fiscal risks report

Climate change

Box 3.1: Accounting for natural capital

As discussed in Chapter 4, there are clear benefits from taking a wider view of the public sector
balance sheet of assets and liabilities when thinking about long-term fiscal sustainability
broader measures like public sector net worth provide a more complete picture than narrow
ones like net debt. But even the broadest of these measures fail to recognise natural assets the
very assets affected by climate change and therefore the costs associated with their depletion.
Asset coverage in economic statistics and commercial accounting is typically limited to those
from which economic or production value can be drawn, and where ownership rights can be
assigned. This means a significant proportion of natural capital (such as air, oceans) is excluded
from balance sheet metrics, while their depletion is also excluded from measures of economic
flows such as GDP.
Recognising this, the ONS publishes the UK Environmental Accounts and the UK Natural Capital
Accounts in addition to the National Accounts on which GDP is based. Both are produced in
Environmental Economic Accounts (SEEA),a which
encompasses a broader asset boundary in physical terms (such as air emissions and water) than
the System of National Accounts (SNA), but are recognised and valued in the same manner as
the SNA.
These frameworks and statistics provide a building block for developing economic accounting
that considers the impacts of climate change. For example, the Environmental Accounts allow for
the assessment of economic activities and household consumption in generating emissions or
tracking government expenditure (and taxes) related to mitigation or prevention activities. The
Natural Capital Accounts are based on the experimental SEEA Ecosystem Accounting framework,
which although still fully aligned with SNA, is more loosely connected to it than the
Environmental Accounts.
While the central framework for Environmental Accounts looks at environmental assets as
individual resources (such as timber, water, soil), the ecosystem framework for Natural Capital
Accounts considers them within ecosystems. Therefore the scope of the latter covers those
services that contribute to the economic benefits measured in the National Accounts as well as
others not accounted for but that relate to the general functioning of ecosystems (measuring
ecosystem conditions, contributions made to society and wellbeing, as well as ecosystem asset
stocks), which addresses the overarching relationship between the economy, society, the
environment, wellbeing and social progress.

review of the economics of biodiversity, led by Professor Sir Partha Dasguptab. His review
nations need to adopt a system of economic accounts that records an inclusive
measure of their wealth , human, and natural
capital. Its conclusions were echoed by calls from G7 finance ministers and central bank
governors for improved corporate financial reporting standards to capture the costs and risks
associated with climate change.c International accounting bodies such as the Financial Stability
Board, IFRS and IPSAS are working on relevant standards for the public and private sector
entities, while bodies such as the UN and OECD are working on standards for national statistics.

Fiscal risks report 90

 Climate change

In its response to the Dasgupta Review, the Government committed to

economic and financial decision-making is geared towards delivering that. d In relation to the
latter, the Treasury and ONS committed to improve their natural capital estimates and examine
the feasibility of developing expanded public sector asset measures, accounting for
environmental assets that yield services (such as carbon sequestration). The Government has
also committed to integrating environmental principles into policy making through a number of
initiatives including regulatory evaluation, cost-benefit analysis, and financing decisions.e
Although broader accounting standards are not yet available, some countries are making
progress in capturing environmental impacts in their policy making processes. Several years ago,

is based on a concept of four capitals: natural capital; human capital; social capital; and
financial and physical capital. Their Budget is presented with a focus on the contribution that
budget policies make to each of these capitals.f
a
System of Environmental-Economic Accounting (SEEA) 2012 Central Framework and SEEA Ecosystem Accounting.
b
Professor Sir Partha Dasgupta, The Economics of Biodiversity: The Dasgupta Review Final report, February 2021.
c
HM Treasury, G7 Finance Ministers and Central Bank Governors Communiqué, June 2021.
d
HM Treasury, The Economics of Biodiversity: The Dasgupta Review Government Response, June 2021.
e
These include: the new Environment Bill; reforming the Better Regulation Framework; Green Book review; funding for the Taskforce
on nature-related Financial Disclosures; and the Green Financing Framework.
f
New Zealand Government, Wellbeing Budget 2021, May 2021.

What can policymakers do to mitigate climate change?

3.17 In light of the market failures and distortions discussed above, what instruments can
policymakers use to overcome them, mitigating climate change by putting economies on a
path to net zero? From the perspective of the fiscal risks they might pose, it is useful to split
them by policy lever:

Carbon taxes. A carbon tax is the most straightforward route to internalise the wider
costs of emissions by placing a uniform price on carbon. The IMF has noted that
the most powerful and efficient, because they allow firms and
households to find the lowest-cost ways of reducing energy use and shifting toward
cleaner alternatives 20 There are several international examples of carbon taxes at the
sector level, but very few apply to almost all emitting sectors.21 South Africa provides
an example of a country that prices carbon solely via a relatively widely-applied
carbon tax, but even this comes with substantial tax free allowances. In the UK, the
carbon price floor only applies to emissions from power generation, but has been
quite successful in reducing emissions in that sector (see paragraph 3.41).

20
IMF, Fiscal Monitor, October 2019: How to Mitigate Climate Change, October 2019.
21
World Bank Group, State and Trends of Carbon Pricing, May 2020.

91 Fiscal risks report

Climate change

Emissions trading schemes (ETS). These provide an alternative to carbon taxes, with the
equally effective if applied to as wide a range of
economic activities
order to deliver a particular emissions path, that can be achieved quite precisely with
an ETS. The EU ETS, which was first introduced in 2005, covers power generation,
energy-
39 per cent of all emissions in 2020. The traded carbon price has varied enormously,

Other tax incentives. Fuel duty in the UK is levied on the use of a fossil fuel, while
vehicle excise duty rates vary by fuel type and (for the first year in which a vehicle is
registered) by emission intensity too. Fuel duty in particular is essentially a carbon tax
on motoring. Landfill tax is levied on waste sent to landfill, thereby in effect taxing the
methane it emits. Other tax-like levers in the UK are the environmental levies that are

technologies. Perversely, these levies incentivise the use of gas over electricity for
household and business customers, thereby slowing the transition to cleaner energy.22

Public spending. As discussed later in the chapter, estimates of the costs of mitigating
emissions include the additional investment and operating costs (and savings)
associated with a wide range of activities that reduce or capture carbon emissions.
Some or all of these could be borne by government. For example, the public sector will
inevitably need to cover costs associated with public buildings and vehicles. But it may
also invest in R&D (either directly or via subsidies to the private sector) or subsidise the
installation of low-carbon technologies like heat pumps in homes or carbon capture
and storage (CCS) facilities.23 Governments might also decide to compensate people
or businesses that lose out from the transition for example, poorer or credit-
constrained families, or emitting sectors of the economy like agriculture that it already
chooses to subsidise. Some or all of these costs might be met by carbon tax revenues.

Regulation and other non-fiscal policies. Fiscal levers can be complemented by non-
fiscal policies that require particular outcomes to be met by particular dates. These are
particularly useful when the desired outcome is for an activity to cease altogether,
which is more efficiently achieved via a ban than a tax. Examples in the UK include the

sale of hybrid cars and vans from 2035,24 and the Future Homes Standard that will
require new homes built from 2025 to be 75 per cent less emitting than homes built
under existing regulations and to be net zero compliant once electricity generation is
decarbonised.25 Unlike a carbon tax, regulations do not yield revenue that can be used

22
As discussed in CCC, Progress in reducing emissions 2021 Report to Parliament, 2021. BEIS, Quarterly Energy Prices, 2020 Annual
Domestic Bills Estimates Supplement, January 2021 shows that in the first half of 2020 industrial customers in the UK faced the highest
electricity prices among EU15 countries but the fourth lowest gas prices, while for domestic consumers in the UK electricity prices were just
above the EU15 average but gas prices were the third lowest.
23
And spending on mitigation comes on top of public spending on adaptation, such as the costs of investment in flood defences.
24
Point 4: Accelerating the shift to zero emission vehicles
ffice, 10 Downing Street, The Ten Point Plan for a Green Industrial Revolution, November 2020.
25
Ministry of Housing, Communities and Local Government, The Future Homes Standard: 2019 Consultation on changes to Part L

Fiscal risks report 92

 Climate change

to meet other costs of the transition, but they can hit existing tax bases, as with the fuel
duty implications of banning new petrol and diesel car sales.

Which policy levers have worked so far to reduce emissions?

3.18 There is a growing body of evidence examining the effectiveness of environmental policies.
Research in this area has faced the challenge of isolating the effects of individual policies,
since they tend not to be introduced one at a time and are confounded by other factors.
And most studies have focused on broader criteria than solely emissions impacts.

3.19 Case studies have therefore reached varying conclusions on the effectiveness of carbon
taxes.26 But several support the argument that countries that deploy carbon taxes tend to
have lower emissions than those that do not, and that, if set at the right level, this instrument
can have a significant effect.27 One case study looked at CO2 emissions in the transport
sector of Sweden, which is subj
emissions were on average 11 per cent lower between 1990 and 2005 relative to the

average) attributed to the carbon tax.28 An OECD summary of recent research suggests that
raising energy prices by 10 per cent (as would be the case with the imposition of a carbon
tax with that impact) would result in a 5 to 10 per cent decline in the use and carbon
intensity of energy.29 One cross-country study cited within this suggests that the EU ETS had
reduced carbon emissions by 10 per cent between its introduction in 2005 and 2012. Some
studies go further and suggest that even a low carbon price can have some impact,
especially when it is known that that it will increase over time.30

3.20 An IMF multi-country review of the impact of a range of environmental policies on the
power sector concludes that both non-market levers (like regulation, emissions limits, and
R&D subsides) and market ones (such as emissions trading and feed-in tariffs) have been
effective.31 These policies are estimated to have contributed to 30 per cent of global clean
energy innovation and 55 per cent of the increase in the share of renewables power
generation. They find less evidence of an impact from a carbon tax in this sector, but note
that the limited take-up of carbon prices globally has held back its effectiveness. (A common
theme across many studies is the importance of pricing carbon correctly to its effectiveness
and that prices may currently be too low. Indeed, one carbon tax that has proved effective is
the carbon price floor in the UK, which has been credited with helping to spur the sharp
reduction in use of coal for power generation.32)

(conservation of fuel and power) and Part F (ventilation) of the Building Regulations for new dwellings. Summary of responses received and
Government response, January 2021.
26
Green, J., Does carbon pricing reduce emissions? A review of ex-post analyses, Environmental Research Letters, March 2021.
27
Best, R., Burke, P. J. and Jotzo, F., Carbon Pricing Efficacy: Cross-Country Evidence, Environmental and Resource Economics, June 2020
November 2019. Also, Green J. provides a table summarising the results of 37 ex-post studies of carbon taxes.
28
Andersson J.J., Carbon Taxes and CO2 Emissions: Sweden as a Case Study, American Economic Journal: Economic Policy, November
2019.
29
OECD, Assessing the Economic Impacts of Environmental Policies: Evidence from a Decade of OECD Research, 2021.
30
Bayer P. and Aklin M., The European Union Emissions Trading System reduced CO2 emissions despite low prices, Proceedings of the
National Academy of Sciences of the USA, April 2020.
31
IMF, World Economic Outlook, Oct The Mitigation Toolkit: How have policies worked so far? .
32
Castegneto Gissey, G., Guo, B., Newbery, D., Lipman, G., Montoya, L., Dodds, P., Grubb, M., Ekins, P., The value of international
electricity trading, University College London and University of Cambridge.

93 Fiscal risks report

Climate change

3.21 One recent cross-country study by Tenreyo and de Silva quantifying the relative impact of
different types of climate-
suggests that carbon taxes and ETSs have been the most effective levers.33 Specifically, it
found that countries with a national carbon tax had emissions 19 per cent less than
countries without one, while the presence of a national ETS reduced emissions by 27 per
cent (considerably greater than the EU ETS impact in the study cited above). By contrast,
emissions were found to decrease by 4 per cent for each additional climate-related law
enacted although it is possible that the aggregate effect of such laws could exceed the
effect of carbon taxes and ETSs given their larger number. This suggests that legal steps can
have an important complementary role alongside carbon taxes (as one would expect when
multiple market failures and distortions are at play). The study found no statistical
relationship between additional climate-related policies (as opposed to laws) and emissions,
which could be because they are typically smaller in scale, or that the impact of effective
policies is balanced out by the lack of impact of ineffective ones.

Chart 3.3: Estimated impact on emissions of the existence of different interventions

30

25
Estimated impact on emissions (per cent)

20

15

10

0
National level ETS National level carbon tax Additional climate-related law
Additional climate-related
policy
Source: Tenreyro, S. and de Silva, T., Climate-Change Pledges, Actions and Progress, 2020, Table 8, Regression 1

Physical, economic, and fiscal risks of global warming

The nature of climate-related fiscal risks to the UK
3.22 The UK accounts for just 1 per cent of global emissions (Chart 3.4), which means the
actions we take to reduce emissions will have little direct influence on global temperatures.
through its

via the positive spillovers that can accrue when its investments in green technologies drive

33
Tenreyro, S. and de Silva, T., Climate-Change Pledges, Actions and Progress, London School of Economics and University of Moratuwa,
October 2020.

Fiscal risks report 94

 Climate change

costs down, thereby incentivising greater deployment and decarbonisation in other countries
(the UK has already contributed significantly to offshore wind technologies, for instance).

Chart 3.4: Global greenhouse gas emissions in 2016: top 20 emitters

China
United States
India
Russia
Indonesia
Brazil
Japan
Iran
EU Germany
Member
UK States Other EU
Canada
Mexico
Saudi Arabia
South Korea
Australia
South Africa
Zambia
Argentina
Nigeria
United Kingdom

Source: Our World in Data

3.23 o bring global

warming under control, we will inevitably be hit by the consequences even if the UK has

transition risks emanating from decarbonising activity at home will be influenced by the
FRRs, physical risks are
largely exogenous, whereas transition risks are more, though not completely endogenous.

3.24 Global developments will not only determine the extent of global warming, they will also
affect the size and frequency of extreme weather events.34 The associated costs represent the
physical risks from climate change. In the absence of policy to mitigate global warming, all
the fiscal risks from climate change would stem from the costs of adapting to these changes.

3.25 These physical risks include risks to existing spending programmes, such as additional
pressures on health systems generated by more intense summer heatwaves (net of reduced
pressures due to less cold winters). And they would include the costs of new programmes,
such as the need to build flood defences as sea levels rise or to install cooling systems in
buildings. Finally, any impacts on the economy, such as higher unemployment or lower
productivity, would feed through to lower tax revenues, thereby increasing borrowing and
public debt or reducing the quantity of public services that could be afforded.

3.26 One US study analysing the potential fiscal costs of unmitigated climate change suggests
that government consumption might increase by 0.32 per cent of GDP for every 1°C

34
Field, C., et al., Managing the risks of extreme events and disasters to advance climate change adaptation, IPCC, 2012.

95 Fiscal risks report

Climate change

temperature rise (which in the US is partly due to more frequent hurricanes and wildfires,
rather than the flooding that is the greater risk in the UK).35 And a 2010 EU analysis
suggested that while the annual direct costs of gradual climate change on a typical member

of GDP) if impacts from

extreme events and indirect effects are taken into account.36

3.27 But in practice the impacts on the economy and public finances may increase with
temperature in a non-linear way, not just because of a higher probability of extreme events
at higher temperatures, but also because of the changing nature of the macroeconomic and
fiscal risks involved. At higher temperatures an increasing number of catastrophic risks
could be faced. Competition for scarce resources could lead to conflict and war, which
could prompt mass migration as habitable land becomes uninhabitable, for instance due to
rising sea levels or desertification. A changing climate will also affect disease patterns, with
mass movements of people potentially fuelling global disease outbreaks, and both the
increase in temperature and new disease patterns precipitating health crises. These factors
could lead to the emergence of energy geopolitics, civil unrest and governance breakdown,
insurance system failures, systemic financial crises, and economic instability.37

Adapting to mitigated climate change in the UK

3.28 Changes to the climate in the UK are significant even under scenarios where emissions are

achieves net zero emissions by 2050, the average surface temperature and the sea level in
the UK will continue to rise until at least the middle of the century ( warmer
and 10 to 30cm higher, respectively, than their 1981 to 2000 averages). Hotter and drier
summers will lead to more frequent heatwaves, droughts and increased wildfire risk (as
experienced in 2020). And warmer, wetter winters will lead to more flooding, and more
extreme storm events (as has also been evident in recent years).

3.29 Adapting the UK economy to these changes that are already in train will require significant
additional action. Every five years, the Climate Change Committee produces a wide-
ranging independent assessment of climate risk in the UK. Its 2021 report details 61 risks
and opportunities for the UK by 2050 due to a changing climate that the Government will
need to respond to in its third Climate Change Risk Assessment (due in 2022).38 It reports
that the gap between the level of risk from climate change and the level of adaptation has
widened in the five years since its previous advice, and that action on adaptation has not
kept pace with climate change: for example, it notes that new-build homes could require
costly future retrofits if, as now, they are not built to address overheating as well as energy
The UK has the capacity and the resources to

35
Barrage, L., The fiscal costs of climate change, AEA papers and proceedings, 2020.
36
Centre for European Policy Studies, The fiscal implications of climate change adaptation: Final Report - Part I, August 2010.
37
See, for example: Raleigh, C., Jordan, L. and Salehyan, I., Assessing the Impact of Climate Change on Migration and Conflict, World
Bank Group, 2008; Sawas A., Workman, M. and Mirumachi, N., Climate change, low-carbon transitions and security, Grantham Institute
Briefing paper No 25, March 2018; the International Military Council on Climate and Security, The World Climate and Security Report
2021, 2021; BIS, The green swan: Central banking and financial stability in the age of climate change, 2020 and Ministry of Defence,
Climate Change and Sustainability Strategic Approach, 2021.
38
CCC, Independent Assessment of UK Climate Risk Advice to Government f , 2021.

Fiscal risks report 96

 Climate change

respond effectively to these [climate-related] risks, yet it has not done so. Acting now will be
cheaper than waiting to deal with the consequences. Government must lead that action 39

3.30 The CCC notes that successfully adapting to the risks from global climate trends will require
adjustments in multiple areas, including:

Land use and soil health. For example, appropriate tree-planting to enhance
biodiversity and minimise species loss, as well as protecting and expanding peatlands
2).

Buildings. Retrofitting houses to prevent them overheating at hotter temperatures and

ensuring new builds are fit for the future climate.

Flood defences. Boosting defences to protect infrastructure and land from rising sea
levels and increasing flooding due to heavy rainfall.

Supply chains and the power system. Investment to mitigate disruptions and make
supply chains resilient to extreme weather events. Making the power system robust to
the variability of renewable energy presents a further challenge.

3.31 It is, of course, very challenging to quantify the costs of adaptation. Indeed, the

to the complexity of calculating costs of adaptation, a national assessment on

adaptation costs in the UK does not exist 40 In preparing its latest independent assessment,
the CCC commissioned a monetary valuation, which put the costs of adapting to different
risks and opportunities in broad ranges. It argued that the benefits of adapting to almost all
climate risks outweighed the costs sometimes significantly.41

An illustration of potential catastrophic risks from unmitigated global warming

3.32 The likelihood of emissions and temperatures rising inexorably has receded progressively as
Governments around the world have taken steps to address climate change and will do so
further if policies are put in place to meet the latest targets. But there remains some value in
illustrating the scale of the potential fiscal risks that the UK might face in such a scenario
given the ebb and flow of collective resolve to tackle global emissions since the turn of the
century and the limited progress made in reducing global emissions to date. Chart 3.5

s) for global emissions scenarios

42
compiled by the IPCC. They show that it is not until later in the century that material
differences emerge between them.

39
adaptation remains the Cinderella of climate change, still sitting in rags by the stove: under-
resourced, underfunded and often ignored Without action on adaptation we will struggle to deliver key Government and
societal goals, including Net Zero itself
40
Department for Business, Energy and Industrial Strategy, Impact Assessment for the sixth carbon budget, April 2021.
41
Paul Watkiss Associates, Monetary Valuation of Risks and Opportunities in CCRA3, 2021.
42
epresentative concentration pathways for atmospheric CO2 concentrations are used as benchmarks for international
climate modelling. RCP 8.5 corresponds to the unmitigated warming scenario shown in Chart 3.1 above. Unfortunately, the other RCPs
do not directly correspond to the NGFS scenarios presented elsewhere in this chapter.

97 Fiscal risks report

Climate change

Chart 3.5: The effect of global emissions paths on UK temperatures

4.5
Unmitigated warming (RCP 8.5)
4.0
Medium-high emissions (RCP 6.0)
Average change in temperature (oC)

3.5
Medium emissions (RCP 4.5)
3.0
Low emissions (RCP 2.6)
2.5

2.0

1.5

1.0

0.5

0.0
2020 2030 2040 2050 2060 2070 2080 2090

Note: Year shown is the mid point of the average temperature taken over 20 years. The low scenario (RCP 2.6) is consistent wi th the
Paris Agreement and implies acheiving global net zero by 2100 - all other scenarios fail to meet Paris Agreement goals.
Source: BEIS, DEFRA, Hadley Centre, Met Office

3.33 Illustrating the fiscal impact of unmitigated climate change requires a departure from our
usual approach to scenario analysis. The past decade demonstrates that the public finances
are all but certain to be subject to significant shocks over a long enough time frame. To
focus on more gradual and predictable pressures on public spending, like those from
population ageing, our long-term projections abstract from such shocks. But given the very
long timescales involved, we have taken a different approach to illustrate the potential fiscal
impacts of accommodating both the increased costs from unmitigated global warming and,
more importantly, the larger and more frequent shocks that it would bring (Chart 3.6).

3.34 The baseline for this illustration assumes the Government balances the current budget,
maintains net public investment at the 2.7 per cent of GDP level reached at the end of our
March 2021 forecast in 2025-26, and adds stock-flow adjustments worth 0.65 per cent of
GDP a year. On that basis, holding all else equal and total borrowing at 2.7 per cent of

based on historical experience in the UK and around the world, layering on the additional
impact of periodic fiscal shocks from recessions and similar events,43 would see debt climb
slowly from around 100 per cent of GDP today to around 170 per cent by the end of the

3.35 It is, of course, difficult to quantify with any confidence the potential long-run economic and
fiscal damage wrought by unmitigated global warming, let alone how that might alter
°C increase in average UK
st
temperatures by the end of the 21 century set out in the most pessimistic Met Office

43
In our 2019 Fiscal risks report, we found there had been seven recessions in the previous 63 years, or one every nine years on average.
International and historical evidence suggests that a typical recession could add around 10 per cent to the debt-to-GDP ratio (see, for
example, IMF, Analyzing and Managing Fiscal Risks Best Practices
per cent of GDP shock every nine years.

Fiscal risks report 98

 Climate change

projection to have severe consequences for the public finances. To illustrate the orders of
magnitude that might be involved, we have assumed that that the cost of adaptation to each
degree of warming raises spending by 0.3 per cent of GDP a year (informed by the
estimates set out above) and that the size and frequency of shocks progressively increases
with rising temperatures to reach twice as large and twice as frequent by the end of the
century (relative to the historical shocks baseline).44

3.36 On these simple, broad-brush assumptions, unmitigated global warming would cause debt
to ratchet up sharply to reach 289 per cent of GDP by the end of the century, as the hit from
each shock increases and the period between them to get debt back down diminishes. At
that point, net debt interest payments might have risen to around 10 per cent of GDP (from
0.9 per cent in 2025-26) and to around 28 per cent of primary revenues (from 2.5 per
cent). And of course these risks would add to, rather than replace, the significant long-term
pressures that we describe in our biennial Fiscal sustainability reports the increased
spending demanded by an ageing society and the non-demographic cost pressures in the
health system pressures that would also be affected, positively or negatively, by such
significant warming (for instance, higher mortality reducing pressures on spending or higher
morbidity due to increased air pollution damaging growth and increasing health spending).

Chart 3.6: Public sector net debt: an illustrative unmitigated global warming scenario
350
Outturn and March 2021 forecast
Stable deficit baseline
300
Historical shocks baseline
Unmitigated global warming scenario
250
Per cent of GDP

200

150

100

50

0
2006-07 2016-17 2026-27 2036-37 2046-47 2056-57 2066-67 2076-77 2086-87 2096-97
Source: ONS, OBR

44
This is an illustrative assumption but is supported by other studies. For example, in Kahn, M., Mohaddes, K., Ng, R., Hashem Pesaran,
M., Raissi, M. and Yang, J-C., Long-Term Macroeconomic Effects of Climate Change: A Cross-Country Analysis, IMF Working Paper
WP/19/215, the authors note that adding the effect of greater climate volatility to their main results that are driven only by the level of
temperatures roughly doubles estimated GDP losses at the global level.

99 Fiscal risks report

Climate change

Costs of decarbonising the UK economy by 2050

3.37 The UK has legislated to reduce net emissions to zero by 2050. This section discusses the
emissions reductions that have been achieved to date and what is left to do by 2050, before
exploring the actions and whole economy costs that might be necessary to achieve that. The
next section then delves into the transitions costs that might be directly borne by the state.

What has the UK achieved so far?

3.38

rvices consumed by
the residents of that country, which adds the greenhouse gases emitted to produce imported

international aviation and shipping emissions. International agreements like the Kyoto
.
International aviation and shipping are also covered in the sixth Carbon Budget and the net
zero target. The UK has done well in reducing territorial emissions since 1990, although up
until the late 2000s this came partly at the expense of higher imported emissions.

3.39
greenhouse gases) fell by 44 per cent (with a sharper, but temporary, pandemic-related
drop last year leaving emissions down 49 per cent on 1990 levels in 2020). Consumption
emissions fell by less, down 29 per cent between 1990 and 2018 (the most recent year for
which data are available).

Chart 3.7: UK greenhouse gas emissions: territorial versus consumption basis

1200

1000
Million tonnes of CO 2 equivalents

800

600

400

200
Consumption emissions
Territorial emissions
0
1990 1995 2000 2005 2010 2015 2020
Source: BEIS, DEFRA

Fiscal risks report 100

 Climate change

3.40 One cross-country study (which looks at just CO2 emissions rather than all greenhouse
gases) estimates that the emissions reduction in the UK since 1990 has been the largest
among the G7 economies and has been faster than the EU average (Chart 3.8).45 The UK
therefore represents a declining share of global emissions.

Chart 3.8: Cross-country territorial CO2 emissions since 1990

40 500
CO2 emissions by country CO2 emissions relative to 1990
400
30 United States
EU27
Billion tonnes emitted

Index 1990 = 100

China
300
India
20 United Kingdom
Rest of world
200

10
100

0 0
1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015
Source: Our World in Data

3.41 The fall in CO2 emissions in the UK mainly reflects lower emissions from power generation
(Chart 3.9, again on a CO2-only basis, which are available by sector). This reduction was
-total
replacement of coal with renewable power sources. In turn, this partly reflects tax and
regulatory interventions that have raised the cost of coal prohibitively (notably the
introduction of the carbon price floor, a carbon tax that overlays the ETS so that power
stations pay a minimum price per tonne of CO246). And it partly reflects sharp falls in the
cost of renewable energy, particularly from wind, thanks to both technological advances
and to further policy interventions, including feed-in tariffs (which subsidise small-scale
-scale generation by
guaranteeing producers a fixed price, with any costs or savings passed to consumers).47

3.42 Lower emissions from businesses have also contributed materially to the economy-wide
reduction in territorial emissions. This reflects both efficiency gains per unit of output in
individual industries, as well as a structural shift in the UK economy away from high-
emissions sectors like heavy industry towards less emissions-intensive activities.48 Reductions
in other sectors have generally been smaller. For example, modest gains in energy
efficiency have lowered residential emissions by 13 per cent between 1990 and 2019; while
improvements in fuel efficiency have offset increases in miles driven to leave road transport
emissions more or less constant (as discussed later in Box 3.2).

45
Global Carbon Project, Global Carbon Budget 2020, 2020.
46
Castegneto Gissey, G., Guo, B., Newbery, D., Lipman, G., Montoya, L., Dodds, P., Grubb, M., Ekins, P., The value of international
electricity trading, University College London and University of Cambridge.
47
See Wind and solar are 30-50% cheaper than thought, admits UK government, Carbon Brief, August 2020, and BEIS, Electricity
generation costs 2020, August 2020.
48
See Reducing UK emissions: progress report to Parliament, CCC, 2020, and 2 emissions have fallen 38%
since 1990, Carbon Brief, February 2019.

101 Fiscal risks report

Climate change

Chart 3.9: Reduction in territorial CO2 emissions between 1990 and 2019
700

600

500
Million tonnes of CO 2

400

300
Other
Buildings
200
Industry and business
Power
100
Transport
2019 levels
0
1990 level Power Industry and Buildings Transport Other Remaining in
Source: BEIS business 2019

3.43 The different paths for territorial versus consumption emissions is partly due to the fact that
although the level of UK manufacturing output in 2019 was roughly the same as in 1990,
the volume of goods consumed in the UK more than doubled, and goods imports almost
trebled, over that period. This raised the emissions-intensity of consumption relative to

per cent between 1990 and 2007, that was more than offset by the emissions embedded in
ion induced
by the financial crisis and have fallen in step with territorial emissions in recent years.

3.44 The pandemic and associated global recession have had an even sharper effect on
emissions than the financial crisis did a decade ago: territorial emissions in the UK dropped
by 11 per cent in 2020, taking them back to a level last seen in the Great Depression of the
1930s (and last seen on a sustained basis in the late nineteenth century). That reflected both
the effects of lower economic activity on energy demand and the limits placed on travel by
stay-at-home advice and social distancing. This was partly offset by energy use in homes,
which increased. As economic and social activity recover, emissions can be expected to
rebound significantly in the near term.49

3.45 Policy interventions have clearly helped in reducing territorial emissions in the UK, with the
carbon price floor cited as an important driver of the drop in emission from power

49
Lockdown measures led to a record decrease in UK emissions in 2020. Most of the falls in sectoral emissions
observed in 2020 are likely to be transient, as they do not reflect structural changes in the underlying economic, social, energy,
transportation or land systems. In the absence of underlying changes, emissions are likely to rebound in most sectors in 2021.
Progress in reducing emissions, 2021 Report to Parliament, 2021.

Fiscal risks report 102

 Climate change

generation.50 But there is considerable uncertainty around the overall effect of policy
measures on economy-wide emissions to date. The OECD maintains an index of the
stringency of environmental policies that summarises how those adopted in the UK might
have contributed to the emissions reductions outlined above (Chart 3.10).51 These have
been tightened over time, including via the introduction of emissions trading, the
strengthening of various regulations in the early 2000s, and the introduction of feed-in
tariffs in the 2010s. On this basis, policies in the UK are now somewhat more stringent than
the OECD average, having been tightened somewhat more quickly over the past decade or
so. According to this metric, less than a third of the overall increase in stringency comes via
revenue-raising measures (taxes and emissions trading), while a fifth comes via feed-in
tariffs that are production subsidies financed by tax-
bills. The remainder of the effect is achieved via standards, regulations, and R&D subsidies.

Chart 3.10: OECD environmental policy stringency index for the UK

4.0

Taxes Trading schemes

Contributions to Environmental Policy Stringecy Index

3.5

Feed-in tariffs Standards

3.0
R&D subsidies Overall UK EPI
2.5
OECD average EPI
2.0

1.5

1.0

0.5

0.0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014
Source: OECD, OBR

What needs to happen to reach net zero by 2050?

Climate Change Committee advice and Government targets
3.46 Having almost halved our emissions since 1990, the Government has legislated to eliminate
any remaining net emissions (i.e. to ensure that gross emissions are offset by gross
removals) by 2050. The 2008 Climate Change Act provides the legislative framework within
hat alongside an overarching target

50
CCC, , 2019.
51
The index is compiled by assigning stringency scores to various different environmental policies based on things like tax rates, carbon
prices or caps set in regulations, which are then aggregated using a weighted average across policy levers. For a full explanation, see
Botta, E. and Koźluk, T., Measuring Environmental Policy Stringency in OECD Countries: A Composite Index Approach, OECD Economics
Department Working Papers No. 1177, 2014.

103 Fiscal risks report

Climate change

within which it will seek to constrain emissions over successive five-year periods. In doing, so
it must take account of the advice of the independent Climate Change Committee (CCC).

3.47
zero, covering the period from 2033 to 2037. In December 2020, the CCC set out its
advice to the Government on setting CB6 to deliver a 78 per cent reduction in emissions of
greenhouse gases relative to 1990 levels by 2035, and demonstrated how its proposals
would be consistent with achieving net zero by 2050.52 The Government accepted that
advice, publishing an Impact Assessment on it in April 2021,53 and draft secondary
legislation has been laid in Parliament that will give CB6 legal force.

3.48 it only includes some

of the short-term fluctuations in emissions associated with the recession and recovery, and it
does not allow for any economic scarring over the long term.54 But over a 30-year horizon,
these omissions are unlikely to change the broad shape of the projections.55 They are
certainly likely to be small relative to the uncertainty around any projection.

3.49 In the remainder of this section we look first at the contribution of different sectors to the

outputs than th

whole economy net costs of achieving that transition as enumerated by the CCC, before
looking more closely at five sectors that contribute the most to the transition to net zero:
surface transport (in particular electric vehicles); buildings (in particular domestic heating);

scenarios to illustrate the uncertainty around these costs. These reflect scenarios in which
societal and behavioural changes, and the pace of innovation, are respectively slower and
faster than in the balanced net zero pathway. Table 3.1 summarises some of the different
assumptions across these scenarios.

52
Climate Change Committee, o, December 2020.
53
Department for Business, Energy and Industrial Strategy, Impact Assessment for the sixth carbon budget, April 2021.
54
-related effects on emissions from aviation and shipping, where the impact was immediately
gradually starts to return to pre-pandemic levels over the next few years
CCC, The Sixth Carbon Budget Methodology Report, 2020.
55
Indeed, in its latest progress report on the path to The temporary fall in emissions in 2020 will have
Progress in
reducing emissions, 2021 Report to Parliament, 2021.

Fiscal risks report 104

 Climate change

Table 3.1:
Tailwinds Balanced net zero pathway Headwinds

100 per cent of sales in 2030 sales in 2032 sales in 2035

Transport
2) 2 HGVs

HGVs
Mixed scenario
outside of industrial clusters 2 H2 for heat
Buildings
2 for heat 2 heat networks
for heat

from renewables from renewables from renewables

Power
H2 use in homes
2 2 (mix of blue 2 (blue) than
Industry and green) and electrification electrification

2, biojet,
H2 production energy-from-waste and
Removals
industrial heat

meat and dairy dairy and 35 per cent meat and dairy
reduction in meat
trees planted by 2035 trees planted
Other trees planted to 2025, 50,000
flying, with 95 per cent use of after 2035 aviation with 20 per cent use
low-carbon fuels of low-carbon fuels
aviation with 25 per cent use
of low-carbon fuels

The path of emissions to net zero

3.50 Although territorial emissions have fallen since 1990, these reductions such as switching
from coal to gas in power generation were relatively easy to deliver. The remaining
emissions will become increasingly difficult to abate because more of the market failures
and other policy challenges described at the start of the chapter apply for example,
technological solutions are more uncertain, while more of the transition must happen in

this, the rate of abatement must rise if the CB6 target and net zero are to be met.

3.51 Chart 3.11 shows the contributions of the four largest emitters and future removals to net

CCC estimates the biggest emitters to be vehicles (23 per cent of the total), buildings (19
per cent, with residential buildings accounting for 15 per cent and non-residential buildings
for 4 per cent), industry (13 per cent) and power generation (10 per cent). By extension,
these four sectors plus the yet-to-be- are also the largest

105 Fiscal risks report

Climate change

sources of future abatement. The remaining gross emissions in 2050 in other sectors are
largely made up of those from agriculture and land use, and from international aviation.

Chart 3.11: CCC balanced pathway for reduction in emissions by sector

600
Other
Removals
500
Power
Million tonnes of CO 2 equivalents

Industry
400
Buildings
300
Vehicles
Net total
200

100

-100
2020 2025 2030 2035 2040 2045 2050
Note: The CCC reports emissions on a slightly different basis to BEIS. Emissions shown here include emissions from international

Source: CCC balanced net zero pathway

The whole economy net cost of the transition to net zero

3.52 -
discounted cumulative costs up to a
given year, with investment costs averaged over the lifetime of the corresponding assets,
including the costs of borrowing to finance those investments). Costs are estimated net of
those that would be incurred in the baseline for example, the cost of installing heat pumps
is net of the cost of replacing gas boilers. For some aspects of the transition, operating costs
are expected to be lower than those of the fossil-fuel technologies that are being replaced,
so they represent a net saving relative to the baseline (this is particularly so for electric cars).

3.53
transition. It is useful for placing a single figure on any chosen scenario for reaching net
-benefit analysis of CB6 and the path to net
zero. In order to generate fiscal scenarios, we need a time profile over the next 30 years as
well as an overall cost, so we use the in-year costs and savings. This is important because
the investment costs are front-loaded, while significant operating savings are projected for
later in the period: for example, the annualised cost in the balanced pathway stands at £16
billion in 2050, compared to an in-year saving of £19 billion in that year (thanks in large
part to cheaper purchase and running costs for vehicles); similarly, the peak annualised cost
is £19 billion and occurs towards the end of the period, in 2047, whereas the peak in-year
cost is £42 billion and occurs much earlier, in 2027. Finally, all costs are presented in real
2019 prices

Fiscal risks report 106

 Climate change

3.54 In the balanced pathway, the CCC estimates the total net cost of abatement across all
sectors of the economy between 2020 to 2050 at £321 billion with £1,312 billion of
investment costs mostly offset by £991 billion of net operating savings. These figures reflect
the whole economy cost of the transition, so exclude transfers between the private and
public sectors (such as fuel duties paid or subsidies received). We discuss the proportion of
the costs and savings that might be borne by the public sector in the next section.

3.55 As noted, net costs peak in 2027, when investment in power generation peaks and
investment in buildings is ramping up. Net costs then fall steadily as operating savings from
improved energy efficiency grow and running costs fall. From 2040 onwards, net operating
savings are projected to outweigh investment costs. And by 2050, the CCC projects a £19
billion annual saving relative to its baseline emissions scenario (Chart 3.12).56

Chart 3.12:
50

40

30

20
£ billion (2019 prices)

10

Vehicles
-10
Buildings
Industry
-20
Power
Removals
-30
Other
Total
-40
2020 2025 2030 2035 2040 2045 2050
Source: CCC balanced net zero pathway

3.56 Over the whole period, the power sector and the buildings sector contribute most to
investment costs (37 and 28 per cent respectively), while vehicles dominate net operating
savings (accounting for 69 per cent of the total savings). The large contributions from power
and buildings reflect different drivers, with power due to the big increase in electricity
generation required to decarbonise energy use in other sectors, whereas decarbonising
buildings must occur at high costs per unit of CO2 abated. Net costs of £321 billion across
the whole economy are more than explained by the power sector alone (£331 billion) and
almost fully explained by the buildings sector (£265 billion). This reflects the large offsetting
effect on net costs from the £352 billion saving in the vehicles sector (Chart 3.13).

56
-associated
costs within the industry, waste and power sectors into the removals sector.

107 Fiscal risks report

Climate change

Chart 3.13: Whole of economy transition costs by sector over 30 years

600

Costs
400

200
£ billion (2019 prices)

-200

-400

-600
Solid columns are investment costs
Shaded columns are operational savings Savings
Diamonds are net costs
-800
Power Buildings Removals Industry Other Vehicles
Source: CCC balanced net zero pathway

Surface transport
3.57 Surface transport is currently the largest emitting sector in the UK, accounting for 22 per

arise from exhaust emissions from road vehicles. The rest is mostly down to rail. Since
1990, surface transport emissions have been largely flat, reflecting offsetting forces:
improvements in new car fuel efficiency have reduced emissions, but that has largely been
offset by mileage, which has risen by 17 per cent (broadly in line with population growth).
The recent popularity of SUVs caused emissions to rise between 2017 and 2019, largely
offsetting the benefit from higher sales of electric vehicles.

3.58 From 2020 to 2050 surface transport makes up 27 per cent of emissions reductions in the

But the reduced running costs of electric vehicles relative to fossil-fuel-powered vehicles,
mean the sector sees a net saving from emissions abatement out to 2050 of £352 billion.

3.59 The transition to net zero is expected to deliver net savings from 2030 onwards as
additional investment is more than offset by lower running costs. Investment costs peak
during the 2020s and are dominated by the purchase of new cars, vans and motorcycles.
Additional investment in new vehicles slows from 2030 onwards (as costs fall), but
investment in infrastructure increases steadily and investment in HGVs picks up, leaving
overall investment costs on a gently rising path. Operating savings rise steadily across the
period to reach a maximum of £30 billion a year in 2050.

Fiscal risks report 108

 Climate change

3.60 Decarbonising surface transport in the balanced pathway reflects a combination of:

Zero-emission vehicles. These account for 80 per cent of emissions abatement in the
sector. Uptake in low-carbon technology is already growing rapidly from a low base

the sales of new petrol and diesel cars and vans by 2030 (see Box 3.2). But electric
vehicles are currently a third more expensive than conventional vehicles. This largely
reflects the high cost of batteries, which account for a third of the total cost.57 Over the
next decade, battery prices are projected to fall rapidly thanks to the technological
advances and economies of scale that come with mass deployment. By 2025, lower
running costs are projected to result in electric vehicles costing less than fossil-fuel
2030 onwards, they are projected to
cost less to purchase too although the effect of
superseded by the forthcoming ban on the sales of new petrol and diesel cars.

Conventional fuel efficiency improvements in new vehicles. Ambitious targets brought

in under EU regulations have led manufacturers to reduce new vehicle emissions.58
Over the next decade, this will continue to help to reduce surface transport emissions.59
But beyond 2030, this channel will no longer operate due to the sale of new fossil-fuel
cars and vans in the UK being banned, with the sale of new hybrids banned from
2035. However, conventional efficiency gains will remain important for reducing HGV
emissions for longer as low-carbon technologies are further from reaching the market.

Behavioural changes reducing transport demand are projected to lower surface

transport emissions by 19 per cent between 2019 and 2050. This can be achieved by
a combination of reducing the need to travel and changing the mode of transport
used. Public transport, walking and cycling offer low or zero-carbon alternatives to
private car travel, for example. Currently, a quarter of mileage is for the purpose of
commuting and, as the pandemic has demonstrated, increased home working can
significantly reduce road usage. Trials have demonstrated that logistical improvements
for freight transport, such as urban consolidation centres (facilities located on the
outskirts of cities that connect long-haul freight with more efficient last-mile deliveries
into a city), can reduce the number of vehicle movements by up to 85 per cent.60

57
Climate Change Committee, The Sixth Carbon Budget. Sector summaries: Surface transport, 2020.
58
See Box 3.1 in our March 2020 Economic and fiscal outlook.
59
In addition to these EU regulations, the Government has committed to introducing post- at least as
ambitious Consultation outcome: CO2 emission performance standards for new
passenger cars and light commercial vehicles, 2020.
60
Transport Systems Catapult, Consolidating public sector logistics operations, 2018.

109 Fiscal risks report

Climate change

Chart 3.14: Surface transport: emissions reductions and whole economy net costs
160 20
Emissions reductions Whole economy net costs
140 10
Million tonnes of CO2 equivalents

120
0

£ billion (2019 prices)

100
-10
80
-20
60
HGVs
Vehicle efficiency -30
40 Rail and public transport
Reduced demand
Cars and vans
Zero-emission vehicles -40
20 Total
Balanced pathway Solid columns are investment costs;
Outturn and baseline shaded columns are operating savings
0 -50
2018 2023 2028 2033 2038 2043 2048 2020 2025 2030 2035 2040 2045 2050
Source: CCC balanded net zero pathway

3.61 There are several key uncertainties around the balanced pathway:

The infrastructure required to enable a smooth transition to zero-emission vehicles and

can currently fill up with petrol or diesel poses a significant challenge. The CCC
estimates that up to 270,000 public charging points will need to be installed before
2030 (up from 25,000 in June 2020), but some estimates are almost double that.61

Around a fifth of the required reduction in emissions in 2050 is assumed to stem from
behavioural changes reducing transport demand. This is despite rising car ownership
and a falling cost of driving, so it is possible other sources of abatement could need to
be greater if mileage continues on the upward trend of recent years.

The projected uptake of electric vehicles is contingent on their affordability and price
relative to conventional vehicles. Sustained reductions in battery prices are assumed
over the next decade following technological advancements, although volatility in raw
materials costs could represent a risk to such projections. Were battery prices to fall by
25 per cent less than assumed, upfront vehicle costs in 2030 would be 6 per cent
higher. Alternatively, if oil prices were lower it would reduce the scope for operating
savings from switching to electric vehicles, lessening the financial incentive to do so.
For example, the CCC has estimated that 150,000 fewer electric vehicles would be
purchased by 2030 if fossil fuel prices were 10 per cent lower.62

3.62 The best option for decarbonising HGVs is not yet clear and the technology is not at present
operational. Over the next decade, the CCC has recommended that the Government
should run large-scale trials to demonstrate industry viability and determine the most
suitable technology. For example, hydrogen fuel-cell vehicles provide one possible solution

61
Transport and Environment, Recharge EU: How many charge points will Europe and its member states need in the 2020s, 2020.
62
Climate Change Committee, The sixth carbon budget. Sector summaries: Surface transport, 2020.

Fiscal risks report 110

 Climate change

for long-range HGVs that appear unsuitable for battery electric solutions, whereas electric
battery solutions could be supported by the use of cateneries.63 To achieve net zero, the
CCC recommends phasing out new HGV sales by 2040. Until then, there is a greater role
for conventional efficiency gains.

Box 3.2: The transition to electric vehicles

Growth in electric and hybrid vehicles
The transition from fossil fuel to electric vehicles is a key el
emissions, accounting for 23 per cent of the total reduction in emissions by 2050. Unlike domestic
heating (Box 3.3), this transition is already well underway. Alternatively-fuelled vehicles made up
more than a fifth of new car sales and around 3 per cent of all cars in 2020. Hybrid cars were still
only 1.1 per cent of new car sales in 2010, but had risen rapidly to reach 14.2 per cent of total
sales last year. Pure electric vehicle sales grew very slowly at first, just reaching 1 per cent of new
car sales in 2019, but leapt to 6.5 per cent of sales last year.a
Because the average car stays on the road for 14 years, the share of electric cars in the total stock
of cars is much lower. By 2025-26, our latest forecast assumes that hybrids will make up 31 per
cent of sales and 9 per cent of all cars, with the corresponding figures for fully electric vehicles
being 16 and 4 per cent (Chart A).b
Chart A: New car registrations and total car stock by power type
140 140
Other
Electric
New car registrations Total registered cars
120 120
Hybrid
Forecast Forecast
Internal combustion engine
100 100

80 80
Per cent

60 60

40 40

20 20

0 0
2010-11 2015-16 2020-21 2025-26 2010-11 2015-16 2020-21 2025-26
Note: Figures before 2015-16 do not include Northern Ireland.
Source: DfT, OBR

Fiscal implications of electric vehicles

The share of electric vehicles in new car sales has important implications for government revenues
from fuel duty and vehicle excise duty (VED), as purely electric vehicles pay neither, and the pace
at which that share has risen has consistently outpaced that assumed in our EFO forecasts (Chart
B). Indeed, if the pace of increase in 2020-21 assumed in our March 2021 forecast of 3.8

63
The Department for Transport is currently funding the designs of trials to test potential solutions to reducing freight emissions. See KTN,
Zero emission road freight funding opportunity, 2021.

111 Fiscal risks report

Climate change

percentage points were to persist over the next five years, rather than the 1.9 percentage points a
year the forecast assumes, fuel duty and VED receipts would be £0.5 billion lower in 2025-26. We
will revisit this assumption ahead of our next forecast.
Chart B: Successive assumptions for electric vehicles as a share of new car sales
16
March 2021
November 2020
14
March 2020
March 2019
12
October 2018
Outturn
10
Estimate
Per cent

0
2017-18 2018-19 2019-20 2020-21 2021-22 2022-23 2023-24 2024-25 2025-26
Note: 2020-21 estimate combines registration data from DfT for 2020 and SMMT for 2021Q1.
Source: DfT, OBR

The sale of new fossil fuel cars will be banned from 2030, with hybrid car sales banned from
2035. On unchanged fuel duty and VED policies, once the entire vehicle stock has turned over,
.
This is a key component of the fiscal cost of getting to net zero emissions.
The role of policy in incentivising the transition

The switch to electric vehicles has so far been slower for private buyers than for businesses which
accounted for two thirds of electric vehicle registrations in 2020.c But public attitudes are changing
a recent Ofgem survey found a quarter of consumers intend to purchase an electric vehicle in
the next five years.d Indeed, following the sales ban announcement in the Prime Mi 10
Point Plan, UK consumer internet searches for electric vehicles doubled overnight.e
Other policy measures can play a role in accelerating this transition in the interim. There is
evidence that countries offering greater financial incentives have higher take-up of electric vehicle,
with the accelerated shift seen across Europe attributed to generous incentives and more stringent
emissions regulations.f,g In the UK, the Government set up the Office for Zero Emission Vehicles in
2009 to support the transition. Since 2011, the Government has utilised a range of policy
incentives, and has put aside funding of £2.8 billion to support the transition, including:
Demand incentives. These include grants that contribute up to £2,500 to the purchase of a
new battery electric vehicle. These subsidies are in place until 2022-23 but have been
successively scaled back since their introduction as vehicle prices have fallen. Demand for

Fiscal risks report 112

 Climate change

hybrids temporarily fell following their exclusion from the scheme in 2018, but sales quickly
picked up again.h The tax system also encourages electric vehicles, as they are exempt
from VED and fuel duty. As well as these financial incentives, the Government, in

the benefits of electric vehicles, such as their lower running costs.

Infrastructure investment. £1.3 billion has been committed to installing public, home and
workplace charging points over the next four years. The number of public charging points
more than trebled between 2016 and 2020 to reach 21,000 in 2020.i Yet this is still less
than 10 per cent of the projected need for public charging points by 2030.

Regulations to discourage use of conventional vehicles. These include the introduction of

emissions-linked charges to drive into some cities by local administrations, such as the
Ultra Low Emission Zone in London from 2019 and Clean Air Zones in Bath and
Birmingham introduced this year. EU legislation has imposed increasingly stringent
emissions targets for car manufacturers since 2015, and these were transferred into UK
law following the end of the Brexit transition period. From 2020, these require stretching
-
r their
average fleet emissions for the purposes of the targets. This is thought to have played a key
role in the uptick in electric vehicle registrations across Europe last year.j
Supply-side measures. These include measures to support technological innovation,
design, and manufacturing of electric vehicles, such as the £30 million of public investment
in battery technology research, as well as proposals for a UK-based lithium extraction
plant. The Government has also stated its ambition to develop a UK Gigafactory producing
electric vehicle batteries at scale.
In contrast to many aspects of the transition to net zero, the need for policy measures to
encourage the switch to electric vehicles is projected to fall away relatively quickly. The CCC
projects that the combined purchase price and lifetime running costs of an electric car will be
lower than for a fossil-fuel one by 2025, and cheaper in terms of purchase price alone from
2030. But much of this depends on further falls in battery prices, which have tumbled over the
past three decades. For example, the current Tesla Model S battery, which costs $13,600, would
have cost more than $500,000 in 1991 a 97 per cent fall in the space of thirty years.k
a
The pandemic meant that 2020 was an unusual year for the car market. Total new car sales fell 29 per cent but electric vehicle sales
were up 184 per cent. The latter in part reflected waiting lists and order backlogs, whereas temporary closures of car salesrooms hit
sales of conventional cars far harder.
b
SMMT, Automotive sustainability report, 2020.
c
SMMT, Business buyers in pole position on Race to Zero as consumers stuck on the grid for electric vehicle adoption, 2021.
d
Ofgem, Consumer Engagement Survey: Insights into consumer attitudes to decarbonisation and future energy solutions, 2021.
e
AutoTrader, Electric engagement surges on Auto Trader as 2030 announcement doubles new car EV leads, 2020.
f
Tietge, U, et. al, Comparison of leading electric vehicle policy and deployment in Europe, 2016.
g
Oxford Economics, The shift to driving on electricity is speeding up, 2021.
h
National Audit Office, Reducing carbon emissions from cars, 2021.
i
ZapMap, EV Charging Stats 2021, 2021.
j
Transport and Environment, Mission (almost) accomplished! Carmakers' race to meet the 2020/21 CO2 targets, and the EU electric
cars market, 2020.
k
Our World in Data, The price of batteries has declined by 97% in the last three decades, 2021.

113 Fiscal risks report

Climate change

Buildings
3.63 The buildings sector is broken down into residential and non-residential buildings. In 2019,
total direct emissions from buildings made up 17 per cent of territorial emissions (largely
from heating homes, with 85 per cent of dwellings using gas central heating), while indirect
emissions (which include electricity use) made up 23 per cent.

3.64
balanced pathway, with an investment cost of £362 billion (28 per cent of the whole
economy total). The net cost of abatement out to 2050 is £265 billion, accounting for
83 per cent of the whole economy net cost of reaching net zero (much higher than its share
in investment costs, thanks to the large vehicles operating savings described above). Most of
these costs are from decarbonising heating in residential buildings (alongside improving
their fabric efficiency through better insulation), as heating is both the largest source of
emissions in buildings and suffers high average carbon abatement costs (see Box 3.3).

3.65 The CCC breaks down the path to reducing emissions from buildings into three main
categories: behavioural change; improvements in energy efficiency; and switching from
fossil fuels to low-carbon alternatives. Over the 2020s, the CCC assumes a large share of

(80 per cent) of abatement eventually

comes from switching to low-carbon fuels, of which 77 per cent in turn represents the
decarbonisation of residential heating. The balanced pathway begins with the phasing out
of the highest-carbon fossil-fuel boilers from 2028, followed by the phasing out of gas
boilers from 2033. No new gas boilers should be installed from 2035, reflecting the roughly
15-year average life of a boiler and the desire to avoid premature scrapping before 2050.

Chart 3.15: Buildings sector: emissions reductions and whole economy net costs
100 25
Emissions reductions Whole economy net costs
90 20
Million tonnes of CO2 equivalents

80
15
70
£ billion (2019 prices)

10
60
Fabric efficiency
50 5
Heat - residential
40
Heat - non-residential 0
30
Other -5
20
Total
10 -10
Solid columns are investment costs;
Baseline shaded columns are operating savings
0 -15
2020 2025 2030 2035 2040 2045 2050 2020 2025 2030 2035 2040 2045 2050
Source: CCC balanced net zero pathway

Fiscal risks report 114

 Climate change

3.66 Some of the important uncertainties around the balanced pathway stem from:

Choices around the technologies to decarbonise domestic heating, and lock-in effects
of past choices. The process of decarbonisation is expected to combine heat pumps,
flexible electric heating, hydrogen-compatible gas boilers and district heating. Under
the balanced pathway, only 11 per cent of homes use hydrogen for heat by 2050, and
the network is fully electrified; in the headwinds scenario, up to 70 per cent of homes
use hydrogen for heat; and in the tailwinds scenario all heating is electrified. This
uncertainty creates a potentially large option value for households in waiting to find
out which technology wins out incentivising continued investment in existing
technologies in the meantime.

The disruptive process of installing heat pumps and improving the heat efficiency of
. Heat pumps are significantly more expensive than existing boiler
technologies. They are also unfamiliar in the UK and require better-insulated homes to
function efficiently. This creates a very large delivery challenge.

Uncertainty over the role of hydrogen. There are uncertainties over how low-carbon
hydrogen production can be, as well as its price and scalability. If it is produced from
natural gas, it will need to be combined with CCS (so-
producing hydrogen this way emits CO2
hydrogen) it will emit only oxygen, but this process requires over twice the energy input
of direct electric heating, and seven times the energy of heat pumps.13 There is further
uncertainty over the safety and feasibility of converting the current gas network to
hydrogen, since hydrogen is more volatile than natural gas (methane) and conversion
of the network would need to be coordinated within local areas. And the future prices
of hydrogen are themselves highly uncertain.64 Hydrogen may also be reserved for
industries that are unable to be electrified, such as aviation, shipping and HGVs.

64
Ueckerdt et al., Potential and risks of hydrogen-based e-fuels in climate change mitigation, 2021.

115 Fiscal risks report

Climate change

Box 3.3: Decarbonising domestic heating: lessons from the switch to natural gas
For the UK, achieving net zero by 2050 in domestic heating is perhaps the greatest challenge of
all the sector transitions. It will require the installation of new equipment and better insulation in
most existing homes, with often high upfront costs and much uncertainty over which technologies

stock has been switched from one form of fuel use to another the decade to 1977 saw 13
million homes converted to use natural gas in a centrally coordinated process.a This box
considers the similarities and differences between that episode and the net zero transition.

(which was produced from either coal or oil) and electric boilers, so that by 1966 gas and
electric boilers combined outnumbered all other domestic heat sources.a This shift was made
possible by the introduction of narrow bore pipes in central heating, which were developed by
the coal industry in the 1950s but were compatible with a range of fuels. Following the discovery
of North Sea gas in 1965, the Government sought to exploit this indigenous fuel supply that was
cleaner and more efficient than either c
imported liquid natural gas) and invested rapidly to convert houses to using natural gas.b
the process
[of conversion] was particularly arduous
c
The process took almost precisely 10
years to complete, with Sir Dennis Rooke, Chairman of the British Gas Corporation, claiming it

completion.b Today, the Climate Change Committee (CCC) estimates that around 28 to 29
million houses will need converting to be net zero compatible, over twice the number.d
Not only was the scale smaller, but the average cost of conversions to natural gas was a fraction
of the cost of switching to zero-carbon domestic heating, even adjusting for inflation and GDP
growth in the intervening decades. The average cost of converting one house in 1966 was
estimated at £30 (around £1,700 in 2019 terms), with the conversion of all 13 million properties
costing just under £400 million (1.0 per cent of GDP in 1966, equivalent to £23 billion in terms
of nominal GDP in 2019).b,e In contrast, the CCC estimates the additional investment in energy
efficiency and heat decarbonisation required to reach net zero in the balanced pathway (in which
heat pumps are the dominant technology deployed) to be approximately £12,000 per house
(around 7 times greater than the natural gas conversions). Combined with more houses to
convert, that gives a total investment of £362 billion in 2019 prices (16 times more than the
natural gas conversion, and equivalent to 17 per cent of 2019 GDP).f,g
While significantly less expensive per houseshold, the conversion to gas heating was still a major
logistical undertaking. Moving 13 million properties to natural gas involved the 12 regional gas
boards, parts of industry (to make new appliances or the parts necessary to convert existing

sell the idea) and the public (to embrace it).b The Government took a central coordinating role,
with a nationalised Gas Council giving the state direct control of the required investment.h

Fiscal risks report 116

 Climate change

The transition to natural gas and to net zero are superficially similar, in that they both involve
millions of private homes converting to a new technology. But in almost every respect, the
transition to net zero is more challenging. It presents an even greater coordination challenge,
covering investment in both heat efficiency and low-carbon technologies, across a larger number
of properties, and over a longer period. This is evident in each aspect of the transitions:
Low-carbon technologies. The switch to natural gas involved a proven technology (natural
gas central heating) with a clear long-term cost advantage over existing heat sources
(coal and oil). By contrast, there is considerable uncertainty around the appropriate
technology for achieving net zero for domestic heating, and thus the option value in
waiting to see how that uncertainty is resolved is high. With net zero there are (at least)
two alternative approaches to decarbonising domestic heating: hydrogen and heat
pumps (both domestic and district). Switching to hydrogen would involve less upfront
investment (attractive for consumers) and would be delivered through the existing gas
network (attractive for existing producersi) and the heating would operate similarly to
existing gas central heating. But there are significant technological and cost uncertainties
around the production and use of hydrogen, and its deployment for use in heating would
require large-scale coordination.j Heat pumps are rare in the UK but are being
increasingly used in several countries across Europe.k But they are more expensive to
purchase and require more heat-efficient buildings to work effectively since they cannot
heat a cold space as quickly as gas central heating. There is also currently a running cost
disincentive to switching to heat pumps because the electricity that they use is more
expensive than gas used by conventional boilers because the cost of environmental levies
is added to electricity bills but not to gas bills.l
Heat efficiency. While there was no heat efficiency element to the natural gas transition,

around 70 per cent have an efficiency rating of EPC D or worse.m,n The Government has
set a target to upgrade existing houses to EPC bands B and C by 2035,n and for all new
builds from 2025 to meet higher efficiency standards according to a new Standard
Assessment Procedure.o To an even greater extent than with the transition to natural gas,
retrofitting and upgrades to existing homes will be invasive and costly to achieve (for
example, requiring the installation of cavity-wall insulation, double/triple glazing, and/or
draught proofing), with estimates for the average cost to upgrade a single home ranging
from approximately £10,000 to £19,000.g The process is likely to be most challenging for
homes with uninsulated solid walls, which make up 28 per cent of homes in the UK.p
Recent experience with Government schemes to promote heat efficiency has not been
encouraging: the 2012 Green Deal was scrapped in 2015 with just 15,000 loans having

Homes Grant was scrapped early with less than 2 per cent of the £1.5 billion earmarked
for the scheme having been allocated in 2020-21, with the Treasury anticipating a further
£295 million (20 per cent) will be allocated during 2021-22.
Workforce and skills. 80 per cent of houses will also need to replace their heating
systems, installing either hydrogen-compatible equipment and/or an electric heat pump.
The Government has a target for residential heat pump installations to reach 600,000 a

117 Fiscal risks report

Climate change

year by 2028,q up from just 27,000 achieved in 2018. If that target were reached
through steady annual rises, converting the remaining 25 million dwellings by 2050
would require around 1 million installations a year from 2029 onwards. The scale of this
investment programme could support a sizeable skilled industry over several decades. For
example, the Construction Industry Training Board estimates that by 2028 an additional
86,000 plumbers will be required, with an extra 350,000 full-time equivalent construction
sector jobs needed overall in the next decade, dropping back to around 200,000
between 2030 and 2050.r This echoes experience in the move to natural gas, where
outside contractors were relied upon, and extensive training was required before they
entered the field 13 dedicated training schools were set up for this purpose.b
The Government is due to publish a new Heat and buildings strategy on the transition soon.
a
Hanmer & Simone, Actors, networks, and translation hubs: gas central heating as a rapid socio-technical transition in the United
Kingdom, 2018.
b
Williams, A History of the British Gas Industry, 1981.
c
Tiratsoo, Natural gas: a study (3d ed.) were using some 35 million
appliances of about 8,500 different types, many of which were obsolete. The number of burners eventually converted was actually
.
d
Element energy, Development of trajectories for residential heat decarbonisation to inform the Sixth Carbon Budget: A study for the
Committee on Climate Change, 2021 and CCC, UK housing: Fit for the Future?, 2019.
e
UKERC, Natural gas network development in the UK (1960-2010): coping with transitional uncertainties and uncertain transitions,
2011.
f
CCC, Carbon Budget 6 Path to Net Zero, 2020.
g
UK Parliament, Environmental Audit Committee, Energy Efficiency of Existing Homes, 2021. There is a range of estimates for costs
to upgrade house energy efficiency, with the upper estimates giving total costs of over £24,000 per dwelling.
h
Jenkins, Government intervention in the British gas industry, 1948 to 1970, 2004.
i
Lowes et al., Heating in Great Britain: An incumbent discourse coalition resists and electrifying future, 2020.
j
Ueckerdt et al., Potential and risks of hydrogen-based e-fuels in climate change mitigation, 2021.
k
IEA, Heat Pumps, 2020.
l
CCC, Progress in reducing emissions, 2021 Report to Parliament, 2021, and Tony Blair Institute for Global Change, £10,000 to
increase your energy bill: making the economics of heat pumps stack up, 2021.
m
MHCLG, Live tables on Energy Performance of Buildings Certificates, last update 14 May 2021.
n
UK Parliament, Energy efficiency: building towards net zero Delivering Residential Energy Efficiency, 2019.
o
MHCLG, The Future Homes Standard, 2019.
p
BEIS, Household Energy Efficiency National Statistics: Detailed Report, 2017.
q
BEIS, Energy White Paper, Powering our Net Zero Future, 2020.
r
CITB, Building Skills for Net Zero, 2021.

Power generation
3.67 Electricity generation accounted for 10 per cent of total UK emissions in 2019. As outlined
above, this is the sector that has been most successfully decarbonised in recent years, with
emissions down 70 per cent since 1990 as the UK switched from coal to gas and
renewables. Decarbonising and expanding electricity generation is central to achieving net
zero, since abatement in most other sectors involves switching from fossil fuels to zero-
carbon electricity as the means of power. This results in power generation being the costliest
sector in the transition to net zero, due to the large expansion of the electricity network.

3.68 From 2020 to 2050 electricity decarbonisation makes up 19 per cent of emissions
81 billion (37 per
cent of the whole economy total). The net cost of abatement out to 2050 is £331 billion,
somewhat more than the whole economy net cost of reaching net zero (as with buildings,
this is higher than its share in investment costs thanks to the net savings from vehicles).
Operating costs become progressively cheaper than the baseline fossil-fuel technologies

Fiscal risks report 118

 Climate change

thanks to the significantly lower cost of running low-carbon technologies (for example, due
to power generation from renewables having no fuel input costs).

3.69

By 2035, fully decarbonising electricity generation by investing in low-cost renewables

and developing markets for gas CCS and hydrogen plants. In power terms, this
represents 400 terawatt hours of new low-carbon generation (compared with 134
terawatt hours generated by renewables in 2020), and a further 50 terawatt hours of
dispatchable, flexible generation to balance fluctuations in generation from renewable
and nuclear sources. As a result, the renewables share of total electricity generation
will need to increase to 70 per cent by 2050 from its 2020 level of 28 per cent.

Also by 2035, phasing out unabated gas, while ensuring security of supply. The CCC
argues that a smooth transition will require the support of a carbon price and/or other
policy mechanism to encourage low-carbon alternatives to fossil-fuel generation, such
as gas CCS, hydrogen production and bioenergy CCS.

Chart 3.16: Electricity supply: emissions reductions and whole economy net costs
120 30
Emissions reductions Whole economy net costs
25
100
Reduction in emissions
Million tonnes of CO2 equivalents

20
Balanced pathway
Baseline
£ billion (2019 prices)

80 15

10
60
5

40 0

-5
20 Investment costs
-10
Operating savings
Net costs
0 -15
2020 2025 2030 2035 2040 2045 2050 2020 2025 2030 2035 2040 2045 2050
Source: CCC balanced net zero pathway

3.70 The main risks with the balanced pathway assumptions for electricity generation include:

Delays to required investment. Power generation projects are typically large and
subject to the time and cost overrun risks that are common to large projects.65

The pace at which renewable generation can be ramped up has implications for other
, where achieving net zero relies on electrification of
existing activities and then powering them with zero-carbon electricity. For example,

65
Indeed, the tendency for cost overruns is sufficiently well-
There is a demonstrated, systematic, tendency for project appraisers to be overly optimistic. To redress
and duration

119 Fiscal risks report

Climate change

electrification contributes 81 per cent of the reduction in emissions in the vehicles

sector in 2050 and 79 per cent of the reduction in the buildings sector.

Space available for wind or solar power generation. In October 2020, the
Government increased its target for offshore wind capacity by 2030 from 30 gigawatts
to 40 gigawatts, for which current leasing of the seabed is believed to be sufficient.66
The balanced pathway requires a further 55 gigawatts of offshore wind capacity. The
required space for this has yet to be identified, meaning the UK will need to hold new
leasing rounds for future developments.67 And the extent to which floating offshore
capacity can remove the space constraint and at what cost is also uncertain.

Industry
3.71 Industry was the third largest source of emissions in 2019, accounting for 12 per cent of
total UK emissions. From 2020 to 2050 decarbonisation in industry contributes 10 per cent
athway, at an investment cost of £46 billion
(4 per cent of the whole economy total). The net cost of abatement out to 2050 is £50
billion (16 per cent of the whole economy cost of reaching net zero). Of the decline in
emissions, 55 per cent is due to switching to less carbon-intensive sources of energy, such
as hydrogen, electricity, and bioenergy. Carbon capture, utilisation, and storage (CCUS)
contributes a further 15 per cent, in part by capturing some of the residual gross emissions
that cannot be reduced through switching to low-carbon fuels. Reductions also come from
improving the energy efficiency of production (notably in the iron and steel, chemicals, and
cement and lime sectors).

3.72 By 2050, the net cost of taking industry to net zero is £61 billion (22 per cent of the total net
cost across all sectors). This consists of £46 billion in investment costs (three quarters of
which are due to switching to low-carbon energy sources) and £15 billion in operating costs
(which unlike most sectors remain higher than in the baseline scenario throughout the
period).68 By 2050, the increase in operating costs is only £1 billion a year. But if costs that
are passed on to downstream sectors via higher prices are included, this figure rises to £2
billion a year. In terms of additional operating costs, switching to hydrogen as a fuel source
costs £1.4 billion, electrification adds a further £900 million, and this is only partly offset by
£1.1 billion worth of operating savings from increased energy efficiency.

3.73 The key elements of the transition to net zero for industry are:

in the early 2020s, improvements in resource and energy efficiency;

from 2025, incentivising industry to switch energy sources (as the costs of switching to
hydrogen combined with CCUS is above those of current power sources); and

investment in CCUS, electrification and hydrogen, which contributes steadily increasing

amounts to emissions abatement between 2025 and 2040, and stabilises thereafter.

66
New plans to make UK world leader in green energy, October 2020.
67
The Crown Estate, The Crown Estate Operational Report, 2019.
68
As discussed below, to present the costs of removals on a more comparable basis to their emissions, we have moved some removals-
related costs from the industry to the removals sector.

Fiscal risks report 120

 Climate change

Chart 3.17: Industry sector: reduction in emissions and whole economy net costs
120 7
Emissions reductions Whole economy costs
6
100 Operating costs
Million tonnes of CO 2 equivalents

Low-carbon energy sources

Low-carbon energy sources
CCUS 5
Efficiency and process Efficiency and process

£ billion (2019 prices)

80 Investment costs
Behaviour 4
Balanced pathway Net cost
Baseline
60 3

2
40
1
20
0

0 -1
2020 2025 2030 2035 2040 2045 2050 2020 2025 2030 2035 2040 2045 2050
Source: CCC balanced net zero pathway

3.74 Key uncertainties around the balanced pathway for industry include:

The pace and scale of switching to low-carbon fuels. Industry is reliant on switching to
greener fuels in order to reach net zero but, unlike in most other sectors, the switch is
projected to raise future operating costs. Without adequate investment in electrified
industrial processes or sufficiently strong incentives to prompt the move to other low-
carbon alternatives, progress could be slower than projected.

Carbon leakage. The combined effect of upfront investment costs and higher future
exposed sectors, as
production is shifted to countries with less stringent emissions policies reducing the
-intensive imports instead.

Removals
3.75 Given the rising cost of eliminating each additional unit of emissions (known as the
gross emissions to

2 is also removed from the atmosphere to bring net emissions down

to zero. The technologies required for these removals are not currently available at the scale
required to achieve net zero indeed removals had no impact on UK territorial emissions in
2019. From 2020 to 2050 removals contribute 7 per cent of emissions reductions in the

economy total). The net cost of the removals sector out to 2050 is projected to total £101
billion thanks to operating costs of £61 billion, thereby accounting for 31 per cent of the
whole economy net cost of reaching net zero.69

69
Of this £101 billion total cost, only the £4 billion costs of direct air capture are recorded by the CCC as a cost of the removals sector,
with the remaining £97 billion being recorded as costs for the sectors that use removals technologies (such as the power stations and
industrial plants that utilise bioenergy with CCS technologies). To present them on a more comparable basis to emissions, we have
therefore moved all removals-related costs into the removals sector.

121 Fiscal risks report

Climate change

3.76 In the balanced pathway, direct air carbon capture and storage (DACCS) is not developed
until 2040, but bioenergy with carbon capture and storage (BECCS) starts having a material
impact on emissions in the 2030s. These work by converting biomass, biogas, and biowaste
into energy and, in the process, capturing the carbon sequestered in them so that it is not
emitted into the atmosphere. (This is achieved by, for instance, extracting hydrogen from
them, burning them in power stations, or using them in industrial processes.)

Chart 3.18: Negative emissions and whole economy costs from the removals sector
10 8
Emissions removed Whole economy net cost
0 7
Other
Million tonnes of CO 2 equivalents

Direct air capture

6
-10 Manufacturing and construction

£ billion (2019 prices)

Fuel supply
5
-20 Operating costs
Total
4
-30
3
-40 Other
Direct air capture 2
Manufacturing and construction
-50 Fuel supply
Power 1
Baseline
-60 Total removed
0
2020 2025 2030 2035 2040 2045 2050 2020 2025 2030 2035 2040 2045 2050
Source: CCC balanced net zero pathway, OBR Source: CCC balanced net zero pathway

3.77 The upfront capital costs from direct capture are supplemented by ongoing operating costs
(i.e., the market price of each tonne of CO2 is unlikely to cover the cost of removing it from
the air), so unless it is entirely paid for by the state, its costs are likely to be passed on to
consumers in the sectors with residual emissions in 2050 (like aviation and agriculture) via
higher prices. If this cost were higher than any carbon tax that was in place, it would be
expected to prompt further changes in behaviour in these sectors.

3.78 Given that removals technologies do not yet exist at the scale required to deliver net zero,
there is a material risk that removals arrive later or turn out significantly more expensive
than expected. This would force a move to more expensive methods of abatement. But it is
also
abatement cost for DACCS is £179 per tonne of CO2 (in 2019 prices), but some estimates
are much lower.70 In essence, the cost of DACCS sets a ceiling on the abatement costs that
should be paid in other sectors, so this hugely uncertain area could be critical to the path
taken elsewhere (for example, in domestic heating, where abatement costs per tonne of
CO2 are high and the volume of emissions to abate is also high).

70
For instance, one study estimates that the costs could eventually fall under £70 a tonne in some scenarios: Keith D., et al., A Process for
capturing CO2 from the atmosphere, 2018.

Fiscal risks report 122

 Climate change

Other sectors
3.79 The remaining sectors are agriculture, aviation, shipping, waste, land use change, F-gases

decarbonisation in these sectors makes up the remaining 23 per cent of the reduction in

of the whole economy total). The transition to net zero by 2050 for these remaining sectors
delivers a net saving of £74 billion thanks to various factors, including increasing reductions
in operating costs in the fuel supply sector, as well as smaller reductions in operating costs
assumed in the aviation and waste sectors.

3.80 These other sectors represent some of the larger sources of residual gross emissions in 2050
in the balanced pathway. Of the 96 million tonnes of residual greenhouse gas emissions
the CCC projects across the whole economy in 2050, agriculture accounts for 37 per cent,
aviation for 24 per cent, and land use, land-use change and forestry for 21 per cent.

Fiscal implications of the transition to net zero emissions

3.81 An assessment of the fiscal risks posed by the transition to net zero must take account of
four different ways in which this transition can affect the public finances:

First, government is likely to be called upon to bear some of the direct cost of transition
described above, at the very least for the buildings it occupies and vehicles it operates.

Second, it faces a direct loss of tax revenues linked to fossil fuels and emissions.

Third, it could derive a direct revenue benefit by taxing carbon more heavily.

Fourth, it must contend with the indirect effects (which could be negative or positive) of
the transition on the public finances via wider economic outcomes.

3.82 In this section we explore the three direct channels. We combine these with a set of
economic scenarios based on those published by the Bank of England to capture the
indirect consequences and generate a set of alternative fiscal scenarios for making the
transition to net zero.71

The share of the costs borne by public spending

3.83 The Government has not taken a view on the extent to which public spending should bear
the cost of the transition to net zero. As such, we need to make some assumptions to
underpin the scenarios we present later in the chapter. The extent to which these
assumptions can be based on firm evidence varies considerably, and this is a key area
where further analysis, and future Government statements, will help us to refine our climate-
related fiscal analysis in the future.

71
Bank of England, Key elements of the 2021 Biennial Exploratory Scenario: Financial risks from climate change, 2021.

123 Fiscal risks report

Climate change

3.84 With those caveats in mind, Table 3.2 summarises our assumptions about the share of costs
borne by government for the next three decades in each sector. The central variant would
result in the state bearing around a quarter of the cost in each decade, as the public shares
of individual lines typically fall over time but the costs become increasingly concentrated in
areas where the public share is relatively high. In the low spending variant, costs are
somewhat less than half those in the central case and the overall share does fall over time.
In the high spending variant, they are somewhat less than twice the central case and are
more uneven across the decades. These assumptions result in overall net impacts on public
spending across the period from 2020 to 2050 of £152 billion in the low variant, £344
billion in the central variant and £553 billion in the high variant (all in real 2019 prices).

3.85 The key assumptions underpinning the figures for the central variants in each sector are:

Vehicles. Policy is reasonably well defined for electric cars and vans given the sales
ban and their falling costs relative to fossil-fuel vehicles. We assume that modest grant
funding for electric cars continues until the mid-2020s, resulting in 20 per cent of
investment costs being borne by the public sector. Investment in car charging
infrastructure is relatively modest at 50 per cent of the total until 2025, and 20 per
cent thereafter (a path that is assumed to cover commercially unviable routes, with
higher spending initially to speed up the transition). For other vehicles (including
railways, buses and HGVs), the shares for both vehicles and infrastructure are
considerably higher. For the purchase of these other vehicles, the public share is
assumed to fall progressively from 100 per cent at the start of the period to 25 per
cent by the end as private investment picks up. For infrastructure, it is 50 per cent
throughout (with the private sector being charged for some of the costs). Finally,
government bears the investment costs of the 3 per cent of all vehicles that are
estimated to be owned by the public sector, while also accruing 3 per cent of the large
operating savings relative to fossil-fuel vehicles.

Buildings. Given large upfront costs in terms of insulation and heating equipment
(which would otherwise be a barrier to many households making the purchase, despite
due to improved efficiency),
some public sector cost in terms of grants and other subsidies seems plausible, but at
what scale is presently unknowable. For simplicity, we assume the state meets all the
costs for residential households in the lowest 15 per cent of the income distribution,
half the costs of the middle 70 per cent of households, but none of the costs for the top
15 per cent of the distribution. This yields an overall figure of 50 per cent of residential
costs that is stable across the whole period. For non-residential buildings, we assume
the government bears the cost for all public buildings (25 per cent of the total by
value), and bears 20 per cent of the costs for private non-residential buildings (for
example, those for which future operating savings would not be sufficient to make the
investment viable in terms of private returns to the building owner). We also assume
that the state will initially meet 90 per cent of the costs of district heating systems that
meet the needs of several buildings (including both public and private owned), but that
this falls to 60 per cent by 2050 as private investment builds up. The public sector also

Fiscal risks report 124

 Climate change

benefits from operating savings as a result of improved energy efficiency of public

buildings (5 per cent of residential and 25 per cent of non-residential buildings).

Power. As is currently the case, costs of low-carbon power generation are assumed to
be largely privately funded thanks to the additional certainty over future returns offered
by contracts for difference (the cost of which are borne by consumers, not
government). But we assume the state initially covers 10 per cent of investment in
strengthening the electricity network, dropping to 5 per cent from 2030 onwards.

Industry. Drawing on the detailed sector-level output assumptions underlying the Bank
ly action scenario described later, we assume that the state initially
meets the full costs for all the industries that see growth reduced by more than 2.5
percentage points relative to the baseline as a result of the burdens imposed to reduce
emissions (which amounts to 60 per cent of the initial costs). In effect this assumes the
state provides sufficient offsetting subsidies to ensure industries do not relocate or
decline more rapidly due to competition from countries with less stringent carbon
abatement policies. Costs are assumed to decline to cover only those industries facing
more than 5 per cent output losses by 2050 (amounting to 25 per cent of costs).

Removals. We assume the state initially meets all the capital and operating costs of
removals, with the share dropping to 50 per cent of both by 2050. The public share is
assumed to be non-zero in 2050 as the state is assumed to continue subsidising some
activities, like agriculture and land use change, which make up 57 per cent of residual
emissions in 2050, and which it currently subsidises in other ways. It also reflects the
fact that emissions from industry account for 25 per cent of capital and 18 per cent of
operating costs of removals in 2050, for which, as described above, 25 per cent falls
to the state. In all, funding 50 per cent of agricultural removals, 100 per cent of land
use change associated removals, and 5 to 6 per cent of industry associated removals,
means around half of all removal costs in 2050 fall to the state.

Other sectors. We assume the state initially meets 80 per cent of costs in other sectors,
comprising all the costs associated with land use and waste, and roughly half the costs
associated with aviation, shipping, agriculture, fuel supply, and F-gases. This share
falls progressively to 50 per cent by 2050, at which point the state only covers costs
related to land use and waste.

3.86
whole economy costs, we have also produced high and low public spending variants for
each assumption. In broad terms, the low variant is intended to represent a lower bound in
which the public sector deals only with its own assets. By contrast, the high variant seeks to
represent the upper end of costs the public sector might plausibly bear, with, for example,
the state taking on almost all infrastructure costs in the vehicles, residential buildings,
industry, and removals sectors. The 27 per cent of the £1,408 billion costs over the whole
period that is borne by the state in the central variant therefore falls to just 13 per cent in the
low variant, but rises to 41 per cent in the high variant.

125 Fiscal risks report

Climate change

Table 3.2: The share of costs borne by public spending

Whole economy Public share of costs (per cent)
cost/saving 2020s 2030s 2040s Total
£ billion (2019 prices) Low Central High Low Central High Low Central High Central
Costs
Vehicles
Cars 213 11 11 20 3 3 13 3 3 3 6
Car infrastructure 35 20 29 70 20 20 60 20 20 50 21
Other vehicles 69 71 85 94 25 62 85 0 39 76 52
Other infrastructure 15 25 50 75 25 50 75 25 50 75 50
Total 332 16 18 28 11 21 38 6 18 33 19
Buildings
Residential 254 7 44 65 7 44 82 7 44 89 44
Non-residential 142 28 43 47 27 42 53 25 43 59 42
Total 396 15 43 58 14 43 72 13 44 79 45
Power 481 4 7 10 0 5 10 0 5 10 6
Industry 46 24 54 89 21 42 77 19 31 66 38
Removals 101 85 89 93 69 75 81 50 59 67 64
Other 52 59 72 84 41 58 75 30 50 65 60
Total costs 1408 15 26 36 12 27 44 12 26 43 27
Savings
Vehicles -684 3 3 3 3 3 3 3 3 3 3
Buildings -131 5 5 5 5 5 5 5 5 5 5
Other -272 1 1 1 1 1 1 1 1 1 1
Total savings -1086 2 2 2 3 3 3 3 3 3 3
Memo: Net cost (£ billion) 321 46 84 115 58 138 228 45 113 195 344

3.87 th these
public shares would see public spending on the transition to net zero ramp up through the
2020s, stabilise at a high level in the 2030s, and then fall back in the 2040s (Chart 3.19).
The majority of the costs to the state come from decarbonising buildings, which costs net
£164 billion over three decades (48 per cent of the total).

Fiscal risks report 126

 Climate change

Chart 3.19: Costs to public sector of the transition to net zero

30
Other High share
Removals Central share
25
Industry Low share
Power
20
£ billion (2019 prices)

Vehicles
Buildings
15

10

0
2020 2025 2030 2035 2040 2045 2050
Source: CCC balanced net zero pathway, OBR

3.88 The Charter for Budget Responsibility requires us to base our analysis on current
government policy where it is stated. In the scenarios presented later in the chapter, we
therefore assume costs to the public sector are covered by the departmental spending totals
up to 2025-26 that were set out in the March Budget, and so do not increase spending
relative to the baseline, which is based on our March EFO forecast. (In effect, this spending
therefore adds to the medium-term spending pressures described in Chapter 2.) From
2026-27 onwards, we assume that the costs are in addition to our assumed baseline. In all
three variants, this means around 10 per cent of the cost across the whole period is treated
as being allocated from within the baseline, while the bulk of it adds to borrowing.

Direct impact on receipts

Receipts losses associated with decarbonisation
3.89 The Government currently taxes carbon indirectly by levying several duties on motoring,
aviation and waste. As these sectors decarbonise, the tax bases underpinning these
revenues will gradually erode. Based on the assumptions set out in th
pathway, overlaid with standard assumptions on the uprating of these duties,72 we estimate
that receipts that are forecast to be worth 1.6 per cent of GDP in 2025-26 could be lost as a
result of decarbonisation by 2050.73 As Chart 3.20 shows, this comprises:

Fuel duties (which are directly levied on the consumption of fossil fuels). These account
for by far the largest share of revenues lost. Revenues that are worth around 1.2 per

72
lished on our
annually alongside the Budget.
73
These estimates are relative to a counterfactual in which the revenues would otherwise have remained constant as a share of nominal
GDP from 2025-26 onwards at the level reached at the end of our March 2021 Economic and fiscal outlook forecast.

127 Fiscal risks report

Climate change

cent of GDP in 2025-26 are halved by the mid 2030s and fall to virtually zero by
2050. The overwhelming driver of this change is the gradual electrification of the road
vehicle fleet. The CCC scenario assumes that sales of new conventional cars, vans and
plug-in hybrids are ended by 2032 at the latest (three years ahead of the date the
Government has stipulated in its sales ban). This leads to the electrification of around
a third of the car and van fleet by 2030, rising to around two-thirds in 2035, almost
90 per cent in 2040 and close to 100 per cent in 2050. The HGV fleet is assumed to
decarbonise almost 97 per cent of the stock by 2050 through the transition to both
battery electric and hydrogen fuel cell vehicles. The CCC also assumes that
behavioural change will reduce mileage, which further weighs on receipts.

Vehicle excise duties (VED) are levied on vehicles using public roads in the UK, but
battery electric vehicles are exempt. VED receipts therefore follow a similar path to fuel
duty, falling to almost zero by 2050 as the fleet is electrified. This accounts for around
0.3 per cent of GDP of the overall tax loss relative to our 2025-26 forecast.

Air passenger duties (APD) are levied on passengers flying from UK airports to
domestic and international destinations. The CCC assumes that carbon abatement is
largely achieved by a reduction in passenger numbers, with only 25 per cent growth by
2050 relative to 65 per cent in the baseline. The fiscal cost of this would be tiny,
lowering APD receipts by 0.07 per cent of GDP relative to our 2025-26 forecast.

Landfill tax and the plastic packaging tax are charged per tonne of waste and
production respectively. The CCC assumes that emissions from the waste sector fall by

prevention, increased recycling and bans on sending biodegradable waste to landfill.

We assume this has a uniform effect on both tax bases, generating large reductions in
small tax bases, thereby lowering receipts by 0.03 per cent of GDP by 2050.

3.90 One conclusion from this a

represent a risk to our most recent medium-term receipts forecast of 0.1 per cent of GDP
(£1.8 billion) in 2025-26. In large part, this reflects our less aggressive assumptions about
the pace at wh

petrol and diesel, then hybrid, car sales). That said, as discussed in Box 3.2, we have
repeatedly revised up the pace of this transition and may need to do so again.

Fiscal risks report 128

 Climate change

Chart 3.20: Loss of motoring, aviation and waste revenues in the balanced pathway
0.2

0.0

-0.2

-0.4
Per cent of GDP

-0.6

-0.8
Landfill and plastics taxes
-1.0
Air passenger duty
-1.2
Vehicle excise duties
-1.4
Fuel duties
-1.6
Total
-1.8
2019-20 2022-23 2025-26 2028-29 2031-32 2034-35 2037-38 2040-41 2043-44 2046-47 2049-50

Source: ONS, OBR

3.91 We have not made any adjustment for North Sea oil and gas revenues. Our latest forecast
assumes that oil and gas production respectively fall by 6 and 7 per cent a year on average
over the medium term. In the longer term, production is expected to continue to decline as
resources are depleted and the cost of extracting them rises. The rate of decline is uncertain
and could be affected by decisions on future licensing rounds that are expected to be

nstallations

Alternate scenarios for receipts losses from decarbonisation

3.92 There is some uncertainty over the speed at which these revenues will be lost, but if the
whole economy net zero emissions target is met, there is little uncertainty over their ultimate
disappearance. We can illustrate the uncertainty around the time profile using the
21):

In the headwinds scenario, revenues are sustained for longer, primarily due to slower
take-up of electric vehicles. Conventional car and van sales are only ended by 2035.
Fuel duty revenues halve a year later than in the balanced pathway.

In the tailwinds scenario, revenues fall faster. Fuel duty revenues halve a year earlier
than the balanced pathway, driven by faster take up of electric vehicles and the ending
of conventional car and van sales by 2030. Behavioural change is also much greater,
with car mileages down by 34 per cent relative to the baseline (compared to 17 per
cent in the balanced pathway). Demand for air travel is assumed to fall by 15 per cent
between 2018 and 2050 (relative to the 25 per cent rise in the balanced pathway).

129 Fiscal risks report

Climate change

Chart 3.21: Motoring, aviation and waste tax revenues under alternative scenarios
2.0

1.8

1.6

1.4
Per cent of GDP

1.2

1.0

0.8
Balanced pathway
0.6
Headwinds scenario
0.4
Tailwinds scenario
0.2
March 2021 forecast
0.0
2019-20 2022-23 2025-26 2028-29 2031-32 2034-35 2037-38 2040-41 2043-44 2046-47 2049-50

Source: ONS, OBR

Receipts gains from levying a carbon tax on all emissions

3.93 The Government currently runs an emissions cap-and-trade scheme, the UK Emission
Trading Scheme (ETS). It was launched earlier this year as the post-Brexit successor to the
EU ETS, in which the UK had previously taken part as a member of the EU. It covers only a
subset of emissions, as is typical for such schemes74 covering power generation, energy-
intensive industry, and commercial aviation. In 2019 the EU ETS covered 25 per cent of UK

electricity generation (in addition to them being covered by the ETS). It is currently set at £18
per tonne of CO2. Extending the coverage of carbon taxation to all emissions, and at a
higher and rising rate per tonne, therefore has the potential to yield significant additional
tax revenue. This could be achieved in different ways for example, by extending coverage
of the UK ETS and pushing the traded price of CO2 higher by auctioning progressively fewer
permits, or by imposing a full carbon tax and raising the tax rate progressively over time.
For simplicity in the rest of the chapter

3.94 To generate illustrative paths for carbon tax revenues in the scenarios presented later in this
chapter, we need to make assumptions for the emissions that will be taxed and the rate at
which they will be taxed. For emissions, we simply use gross emissions
balanced pathway (i.e. excluding the negative emissions from removals and land sinks). For
the tax rate, in the absence of any Government statements about use of a full carbon tax,
we combine elements from Bank of England and CCC scenarios:

74
As of 2020, 22 per cent of emissions around the world were covered by a carbon pricing system, either a carbon tax or a cap and
trade system. See World Bank, State and Trends of Carbon Pricing 2020, 2020.

Fiscal risks report 130

 Climate change

. The Bank of England scenarios that form the basis of

our net-zero fiscal scenarios (described from paragraph 3.99) assume a rising path for
the shadow carbon price. This price covers both the explicit price set by a carbon tax
and the implicit or shadow price that captures the tax-equivalent effects of other
mitigation policies like bans on certain activities and other regulatory interventions. It
rises over time reflecting the need to incentivise progressively more costly forms of
carbon abatement as the lower-hanging fruit are picked off.75

The proportion of the shadow price delivered by a carbon tax. For simplicity, the Bank
has assumed that half the shadow price is delivered via a carbon tax. We use that as
the starting point for the tax rate in our scenarios. In the longer term, the amount
raised in tax per tonne would be unlikely to exceed the cost of removals via direct air
CCS (which in effect puts a ceiling on marginal abatement costs, as firms could, in
theory, choose either to pay for the cost of removal or pay the tax). So to keep the tax
ed pathway, we assume the
proportion of the shadow carbon price delivered by taxing emissions falls to a quarter
in 2050-51. This would be consistent with a rising share of the shadow price being
delivered by non-tax policies such as outright bans.

The resulting carbon tax rate. The assumptions described above yield a tax rate (in real
2019 prices) that starts at £101 per tonne in 2026-27 (the first year beyond our March
EFO forecast horizon). That would be significantly higher than the rates underpinning
our March forecast, which were based on EU ETS carbon price futures and existing
policy with respect to the carbon price floor (five times higher than the ETS alone and
three times higher than the ETS plus carbon price floor). The tax rate then rises steadily
further to reach £187 per tonne at the scenario horizon in 2050-51 (Chart 3.22).

Chart 3.22: Real-terms carbon tax rates: outturn and scenario assumption
220
Carbon tax rate: early action scenario
200
UK carbon price outturn and forecast: ETS
180
UK carbon price outturn and forecast: ETS plus carbon price floor
£ per tonne of CO 2 (2019 prices)

160

140

120

100

80

60

40

20

0
2015-16 2018-19 2021-22 2024-25 2027-28 2030-31 2033-34 2036-37 2039-40 2042-43 2045-46 2048-49
Source: Bank of England, Datastream, HMRC, OBR

75
See, for example, Burke, J., Byrnes, R. and Fankhauser, S., How to price carbon to reach net-zero emissions in the UK, 2019.

131 Fiscal risks report

Climate change

3.95 We base the first five years of our scenario on tax policy as set by the Government at the
time of our March forecast, so assume for these projections that all emissions become
subject to carbon taxes at higher rates from 2026-27 onwards. (In effect, this means that

policies in the first five years of the scenario.) On this basis, additional carbon tax revenues
start at 1.8 per cent of GDP in 2026-27 (a step change that in reality one might expect to
be phased in over time). Thereafter revenues fall steadily to reach 0.5 per cent of GDP in
2050-51, as falling emissions more than outweigh the positive effect of the continuously
rising tax rate. These additional revenues come from two sources:

Sectors that already pay a carbon price under existing policy (electricity supply, industry
and aviation). Less than 10 per cent of the additional revenue comes from levying a
higher carbon price on these sectors. This raises 0.3 per cent of GDP in 2026-27 and
falls to less than 0.1 per cent from 2035-36 onwards as emissions decrease, primarily
in the electricity supply and industry sectors.

Sectors that currently do not pay a carbon price. Over 90 per cent of the additional
revenue comes from expanding the tax base to cover all other emissions. This raises
1.5 per cent of GDP in 2026-27, falling to 0.5 per cent of GDP in 2050-51 as
emissions fall at a faster rate than the tax rate rises.

Chart 3.23: Additional carbon tax revenues

2.0
Currently traded sectors
1.8
All other sectors
1.6

1.4
Per cent of GDP

1.2

1.0

0.8

0.6

0.4

0.2

0.0
2019-20 2022-23 2025-26 2028-29 2031-32 2034-35 2037-38 2040-41 2043-44 2046-47 2049-50

Source: Bank of England, CCC, OBR

3.96 Carbon tax revenues would be sensitive to the pace of decarbonisation and the choice of
tax rate. The effect of differences in either would be linear with the response to differences
in emissions being one-for-one, while the response to a higher tax rate would less than one-
for-one due to its effect on emissions. The risks posed by these sensitivities could be
expected to rise over time. In 2026-27, carbon tax revenues would be relatively certain as

Fiscal risks report 132

 Climate change

they would be spread across many sectors of the economy. By contrast, in 2050-51 around
80 per cent of revenues come from the agriculture, aviation and land-use sectors in which
decarbonisation of activity is least successful. Revenues would be highly uncertain at this
stage since the narrow tax base and high tax rate would mean even small differences in the
pace of emissions abatement could have large effects on revenue.

Total direct impact on receipts

3.97 Combining the receipts lost due to emissions reductions and those gained from taxing all
emissions delivers a net increase in revenue from 2026-27 through to 2035-36 (peaking at
1.6 per cent GDP in 2026-27), but a net reduction thereafter (eventually reaching 1.1 per
cent of GDP in 2050-51) (Chart 3.24). Over the full period, the early gains are ultimately
outweighed by the subsequent losses and in steady-state they would leave the public
finances in a weaker position in the absence of offsetting policy measures.

Chart 3.24: Total direct impact of the transition to net zero on receipts
2.0

1.5

1.0

0.5
Per cent of GDP

0.0

-0.5
Additional carbon tax revenues
-1.0
Receipts lost to decarbonisation
-1.5
Total
-2.0
2019-20 2022-23 2025-26 2028-29 2031-32 2034-35 2037-38 2040-41 2043-44 2046-47 2049-50

Source: OBR

Fiscal scenarios for achieving net zero emissions

3.98 Fiscal risks (both positive and negative) come not only from adapting to whatever global
warming is allowed to occur, but also from our efforts to prevent further climate change.
Other than for the very largest emitters, these efforts will be small relative to the sum of
global mitigation activity. So while the UK Government has considerable control over the
effort that is put into decarbonising activity at home, the fiscal risks posed by the transition

to get to net zero and how low-emissions technologies and their costs evolve.

133 Fiscal risks report

Climate change

3.99 In this section, we begin by exploring the economic and fiscal implications for the UK of a

constructed by the Bank of England for the purpose of exploring its implications for the
financial system.76 77

in which global carbon prices are raised progressively over the next three decades leading
to global CO2 emissions being reduced progressively to net zero by 2050, and is consistent
with limiting warming to 1.8°C by that point. Within this global path, the UK is also assumed
to eliminate all net greenhouse gas emissions by the middle of the century.

3.100 This scenario is consistent with the world and the UK putting in place the policies necessary
to achieve ambitious emissions targets, and doing so in a timely manner, thereby meeting
the objectives of the Paris agreement with a minimum of disruption to economic activity. It is
therefore in some senses an optimistic scenario, since such policies are largely yet to be
set.78 On existing policies alone, much greater warming is in prospect, as outlined in
paragraphs 3.8 to 3.9, and 3.32 above. It is by no means clear yet that net zero will be
achieved.

3.101 After detailing the economic and fiscal implications of this early action scenario, this section
then explores the sensitivity of our results to varying key assumptions in order to illustrate
some of the trade-offs and uncertainties inherent in the transition to net zero. We consider:

action versus inaction in tackling climate change;

early action versus late action in the transition to net zero;

the uncertain implications for productivity of the substantial amounts of investment

necessary to bring emissions down to net zero;

high versus low public spending shares of this investment; and

alternative ways of calibrating long-term policy assumptions in respect of the tax

burden on motoring and the degree to which public investment in decarbonisation is
additional or is allocated from the amounts implicit in our baseline assumptions.

3.102 There are, of course, many other sources of uncertainty that could be explored in future
work. For example, increased whole economy investment could have implications for
interest rates, which, as we document in Chapter 4, the public finances have become more
sensitive to in recent years. But a s
fluctuations in oil prices, which have been the source of macroeconomic shocks in the past.

76
Bank of England, Key elements of the 2021 Biennial Exploratory Scenario: Financial risks from climate change, June 2021.
77
NGFS Climate Scenarios
for central banks and supervisors, June 2021.
78
See Climate Action Tracker, Climate summit momentum May 2021, 2021 and CCC, ribution to stopping
global warming, May 2019.

Fiscal risks report 134

 Climate change

Early action scenario

Climate and economic assumptions
3.103 early action scenario, with the assumed
79
and the path for emissions drawn from the corresponding NGFS
scenario. The shadow carbon price rises steadily over the next 30 years from $30 a tonne to
$889 a tonne (in real 2010 US dollars), such that global CO2 emissions drop to close to net
zero by 2050.80 This is assumed to keep global warming to 1.8°C above pre-industrial times
end of the century.

Chart 3.25: Scenario assumptions: global carbon price, emissions and temperature
900 45 1.9
UK carbon price Global emissions Temperature change

Change relative to pre-industrial levels

800 40 1.8

700 35 1.7
Giga tonnes, CO 2

600 30 1.6

(degrees celcius)
$, 2010 prices

500 25 1.5

400 20 1.4

300 15 1.3

200 10 1.2

100 5 1.1

0 0 1.0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
Source: Bank of England, NGFS Climate Scenarios Database

3.104
baseline without further global warming, but also with none of the costs of achieving that.
So the baseline is empha
degree of warming and all the economic and fiscal costs that that would bring.

3.105 Real GDP ends up 1.4 per cent below the baseline by 2050, with that loss having been
incurred by 2030 and with a peak loss of around 2 per cent in the mid-2030s. These are
modest differences when set against expected growth over the period (Chart 3.26). Higher
carbon prices and other policies at the global level are introduced early and gradually over
the next 30 years, so that the world transitions smoothly to net zero over that period.
Consumers, businesses and financial markets gradually align their activities to a low-carbon
economy, so there is only a moderate crystallisation of transition risks. Successfully
stabilising global temperatures means that physical risks remain limited too, though they do
increase from current levels due to the additional warming that takes place.

79
s the tax-
equivalent effects of other mitigation policies like bans on certain activities and other regulatory interventions.
80
As they are producing a globally consistent scenario for emissions, the NGFS do not provide a path for emissions in the UK. Also, the
global reductions assumed by the NGFS are slightly more frontloaded than assumed in the UK in the CCC s balanced net zero pathway
(although as both scenarios assume emissions fall from current levels to zero over 30 years the average pace is broadly consistent).

135 Fiscal risks report

Climate change

Chart 3.26: Early action scenario: real GDP

160
Baseline
150
Early action scenario
140
Outturn
130
Index 2019-20 = 100

120

110

100

90

80

70
2000-01 2005-06 2010-11 2015-16 2020-21 2025-26 2030-31 2035-36 2040-41 2045-46 2050-51
Source: ONS, OBR

3.106 To construct a fiscal scenario from the starting point of these Bank assumptions, we take the
percentage shortfall in real GDP and apply it to the baseline path for nominal GDP set out
in our latest long-term economic determinants.81 The scenario uses a path for long-term
interest rates that converges to our long-term economic determinants.82

Revenue and spending assumptions

3.107 The fiscal scenarios are constructed using differences from a baseline path that is identical
to our March 2021 EFO forecast until 2025-26. Beyond that point, the baseline assumes net
investment is held at its 2025-26 level as a share of GDP, the current budget is held in
balance, and stock-flow adjustments (loans and other financial transactions that affect debt
but not the deficit) are stable at less than 1 per cent of GDP. This baseline abstracts from the
large long-term fiscal pressures that feature in our Fiscal sustainability reports, although of
course climate change will need to be addressed alongside those pressures, not instead of
them. The baseline therefore assumes that public sector net borrowing (PSNB) is constant at
2.7 per cent of GDP and that public sector net debt (PSND) falls gently from 104 per cent of
GDP in 2025-26 to reach 95 per cent of GDP in 2050-51.

3.108 Constructing the fiscal scenarios involves three steps. First, we add the direct fiscal costs of
the transition to the baseline using assumptions detailed in the preceding section:

Net zero public spending. The direct effects of the transition on public spending are

variant for the public spending share of those costs (shown in Chart 3.19 above).

81
As published as a supplementary release to our March 2021 Economic and fiscal outlook, available on our website.
82
Specifically, it uses the same

Fiscal risks report 136

 Climate change

Net zero receipts losses. These losses of existing tax revenues that are somehow linked

modelling of their implications (shown in Chart 3.24).

Additional carbon tax revenues

scenarios (the tax rate and revenues are shown in Charts 3.22 and 3.23 respectively).

3.109 Next, we add the indirect fiscal consequences of the different path for real GDP. We have
used a very simple approach to model these effects:

Non-climate-related receipts are assumed to move one-for-one with nominal GDP, so

the non-climate-related receipts-to-GDP ratio is invariant between scenarios. This is
similar to the assumption we use in our long-term fiscal sustainability analysis. It is also
consistent with historical evidence that the fiscal drag inherent in the tax system, which
means that, all else equal, receipts move by more than one-for-one with GDP in the
short term, has historically been offset by the consequences of tax policy changes (or
other factors weighing on tax receipts) over the longer term.83

For non-climate-related public spending, we assume that the volume of public services
and public investment is held constant, while all other spending moves in line with
nominal GDP. This is a bespoke assumption for the purposes of these climate
scenarios that illustrates the extent to which spending would rise (or fall) as a share of
GDP to maintain the volume of public services and investment in the baseline. It differs
from our standard long-term approach of assuming that all spending moves one-for-
one with GDP, which in effect assumes that policy settings are adjusted to reflect the
amount of revenue the economy can generate. This results in an elasticity of a half on
total public spending relative to GDP, since public services and investment spending
make up roughly half of total spending. This means that the spending-to-GDP ratio
varies inversely with differences in the real GDP path across the scenario.

3.110 Finally, we add the debt interest consequences of any differences in borrowing (positive or
negative) between the scenario and the baseline.

Fiscal scenario results

3.111 The fiscal implications of the early action scenario are modest in the first half of the period,
thanks to the largely offsetting effects on receipts and spending, but they become
increasingly negative in the second half of the period (Chart 3.27). This path reflects:

Receipts. Existing emissions-related receipts (mainly fuel duty) fall below the baseline,
but additional carbon tax revenues more than offset these losses initially. This leaves
receipts higher as a share of GDP until 2035-36. But the combination of declining
carbon tax revenues (as emissions fall) and the continuing falls in emissions-related tax
bases means that by 2050-51 receipts are 1.1 per cent of GDP lower.

83
See Table 1 in Belinga, V., Benedek, D., de Mooij, R. and Norregaard, J., Tax buoyancy in OECD countries, IMF Working Papers No
14/110, International Monetary Fund, June 2014.

137 Fiscal risks report

Climate change

Spending. Public spending is higher than the baseline throughout, peaking at 0.9 per
cent of GDP higher in 2035-36 when investment costs and indirect effects of the real
GDP loss are both near their peaks. By 2050-51, spending is 0.5 per cent of GDP
higher. Of this difference, around two-fifths reflects mitigation-related public spending
and three-fifths comes from the smaller economy requiring higher public spending as
a share of GDP to maintain a constant volume of public services and investment.

Borrowing. Higher spending on decarbonisation and the loss of some existing tax
receipts initially raises borrowing, then briefly falls below the baseline between 2026-
27 and 2031-32 thanks to additional carbon tax revenues exceeding the various fiscal
costs. The adverse effect on borrowing then becomes progressively larger as the
receipts consequences of the transition grow, leaving borrowing 2.3 per cent of GDP
higher than the baseline in 2050-51.

Debt. The cumulative effect of this path for borrowing on the debt-to-GDP ratio leaves
it close to the baseline in the first half of the period, but places it on a rising path
thereafter. By 2050-51, PSND is 21 per cent of GDP higher than the baseline
(somewhat less than the overall increase as a result of the pandemic).

Chart 3.27: Early action scenario: key fiscal aggregates

54 54
Non-interest receipts Non-interest spending
52 52
Baseline 50
50
Early action scenario
48 48
Outturn
Per cent of GDP

Per cent of GDP

46 46

44 44

42 42

40 40
38 38
36 36
34 34
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
20 125
Borrowing Debt
115

15 105
95
Per cent of GDP

Per cent of GDP

10 85
75
5 65
55
0 45
35
-5 25
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
Source: ONS, OBR

Fiscal risks report 138

 Climate change

3.112 Chart 3.28 breaks down the contributions from various sources to the difference between
the debt-to-GDP ratio in the early action scenario and the baseline. It shows how the
variation across time is dominated by tax-related assumptions, as emissions-related
revenues fall away progressively while carbon tax revenues initially step up sharply before
declining progressively too. The direct public spending consequences of the transition build
up steadily, as do the indirect effects on public spending as the volume of public services
and non-climate-related investment is maintained in a smaller economy. The debt interest
consequences of all these factors combined build slowly initially, but faster later.

3.113 By 2050-51, the cumulative contribution of higher carbon tax revenues (14.2 per cent of
GDP) offsets around three-quarters of the cumulative loss of existing receipts due to
decarbonisation (19.4 per cent), so these receipts in total explain a quarter of the 21 per
cent of GDP increase in debt (adding 5.2 per cent of GDP). Public spending on
decarbonisation itself explains almost a third of the rise (6.0 per cent of GDP), while
maintaining public services in a smaller economy explains around a quarter of it (4.7 per
cent of GDP). The remainder is largely higher debt interest spending (3.5 per cent of GDP),
with a smaller GDP denominator also raising the debt-to-GDP ratio slightly (1.4 per cent). It
is notable that additional carbon tax revenues over the period as a whole are greater than
the 10.7 per cent of GDP cost of net zero public spending and the smaller economy
combined.

Chart 3.28: Early action scenario: differences from the baseline debt-to-GDP ratio
40
Debt interest spending
Indirect effects
30
Net zero public spending
Net zero receipts losses
20
Denominator effect
Per cent of GDP

Carbon tax revenues

10
Overall effect

-10

-20
2020-21 2023-24 2026-27 2029-30 2032-33 2035-36 2038-39 2041-42 2044-45 2047-48 2050-51
Source: OBR

3.114 This early action scenario provides a plausible net-zero consistent reference scenario against
which to test sensitivities to different assumptions. In what follows, we use the same
economic and fiscal assumptions as in the early action scenario unless otherwise stated.

139 Fiscal risks report

Climate change

Action versus inaction in tackling climate change

3.115 Illustrating the fiscal costs to the UK of global inaction in tackling climate change relative to
our 30-year early action scenario is not straightforward. The costs of mitigation show up
within this horizon, but the benefits of preventing warming largely do not, even though
choices taken by the middle of the century will have lasting influences on the trajectory of
global warming beyond that point. In addition, severe climate change does not just alter a
given central outcome in any given year, it is also likely to increase the likelihood and
magnitude of tail-risk events occurring, due to extreme weather events here and/or the
resulting mass migration and conflicts that might result in hotter, drier parts of the world.

3.116
that provides some insight into this question. In this scenario, it assumes that no action is
taken to tackle climate change beyond the policies already in place before 2021. To
highlight the potential physical risks that would eventually crystallise in this scenario,
significant warming is assumed to take place immediately and reach 3.3°C by 2050. This is
outside the temperature changes implied by most climate models, reflecting a 30-year
shifting forward in time

3.117
action scenario). This means there are both greater pressures (from rising temperatures and
sea levels) and also larger and more frequent shocks (such as heatwaves, wildfires and
severe flooding). As a result, although transition risks are lower than in the early action
scenario, the crystallisation of more severe physical risks lowers the level of GDP to around
8 per cent below the baseline in 2050 (more than five times greater than in the early action
scenario). Moreover, inaction is assumed to result in permanently lower growth, so the GDP
shortfall would continue increasing thereafter.

3.118
not produced a corresponding fiscal scenario. Instead, earlier in the chapter we showed the
illustrative fiscal consequences of completely unmitigated warming in which atmospheric
CO2
result in both higher average temperatures and a much greater risk of crossing climate
tipping points that lead to even more extreme outcomes. (Of course, such a scenario now
appears increasingly unlikely it would fail to take into account mitigation policies already

3.119 Our illustration of the potential costs of completely unmitigated global warming was shown
in Chart 3.6 above, reflecting not only increased pressures on the public finances, but also
that by 2100 shocks occur twice as often as historically and are twice as costly when they
do. This increases debt by over 100 per cent of GDP by 2100, relative to a stylised baseline
without climate change. These figures are based on extremely broad-brush assumptions,
but do serve to highlight the magnitude of the fiscal costs that might be avoided by
successfully stabilising global temperatures in line with the Paris targets.

Fiscal risks report 140

 Climate change

3.120
policies that have already been announced would therefore lie somewhere between this
illustration of catastrophic unmitigated warming and the early action scenario. Introducing
the policies necessary to meet the climate targets that are set out in legislation in the UK and

from the catastrophic scenario. But this is very much still work in progress.

Early versus late action in tackling climate change

The costs and benefits of early and late action
3.121 When considering the fiscal implications of net-zero transition risks it is useful to consider
how the potential impacts of early action might compare to acting later:

Early action. The costs of new technologies tend to fall rapidly once deployment has
reached critical mass, leading to changes in expectations that drive increased demand
and the pressure for innovation. Acting early could place the economy on a better path
than will be available later, for example by increasing productivity and capturing
global market share by being the originator of successful low-carbon technologies, or
from lower input costs associated with being an early adopter giving time for supply
chains to develop and the costs of technology to fall (particularly for domestically
focused markets, where the scope for relying on other countries is more limited). And
some elements of the transition involve very large volumes of activity such as
electrifying thousands of miles of railway or the heating systems in millions of private
homes. In these areas, early and sustained action can mean less pressure on supply
chains and costs than would happen in a later and hurried transition. But these
potential gains need to be weighed against the uncertainties that come with early
action, such as the possibility of investing in what proves to be the wrong technologies,
or of a larger share of investment taking place while unit costs are high.

Delayed action
investments in new technologies, so that they bear the costs of initial uncertainties and
of any trial and error along the way, after which a successful technology can be
adopted at lower cost at a later date. But it also poses risks, with continued investment
in emissions-intensive assets in the meantime increasing the amount of premature
scrapping of those assets in a later and faster transition. A greater share of low-
emissions technologies might also have to be imported, with foreign direct investment
in net-zero industries having flowed to other markets. And if delayed action in the UK
were to take place in the context of early climate action by our major trading partners,
there would be an additional risk of losing export market share if they chose to impose
carbon tariffs at the border.

3.122 One cannot be certain as to which path poses the greatest fiscal risks or opportunities given
the many global and domestic factors at play. Either could be delivered in more predictable
or disruptive ways, with respectively smaller and larger economic costs. One thing is clear
developments since we last looked at climate-related risks in our 2019 FRR point to inaction

141 Fiscal risks report

Climate change

in the race to net zero increasing the risk of being an outlier in the global transition. Not
only have 131 countries now committed to net zero emissions targets or have them under
consideration, those countries now include each of the top three emitters: the United States
(committed to net zero by 2050),84 China (by 2060, peaking by 2030),85 and the European
Union (by 2050).86 Together they made up 49 per cent of global GDP in 2019, before the

momentum is also apparent in the actions of investors and courts in bolstering the emissions
major oil and gas companies in the light of more
87
ambitious official targets. Moreover, the EU is already at an advanced stage in its
consideration of a carbon border adjustment mechanism that one study estimates could, in
affect up to a third of UK exports to the EU, with the steel
sector particularly hard hit due to the high share of its output that is sold to the EU.88

A late action scenario

3.123 rio helps
provide a guide to the potential costs of acting late and in particular of doing so abruptly,
preventing households, businesses and financial markets from adjusting their activities
gradually. In this scenario, the UK and the rest of the world do not take decisive steps to cut
emissions until the 2030s and subsequent action must be more aggressive as a result to
bring global emissions down more sharply to get to net zero by 2050. The sharp rise in the
carbon price necessary to achieve this results in a period of macroeconomic disruption and
leads to carbon-intensive capital being scrapped before the end of its economic life.
Physical risks increase more rapidly than in the early action scenario, but remain modest
relative to an unmitigated climate change scenario. In this scenario, real GDP settles 3.2 per
cent below the early action scenario in 2050, following a period of disruption in which the
loss of output relative to this scenario peaks at 6.6 per cent in 2033-34.

3.124 In addition to this different GDP path, we vary our assumptions about tax and spending:

We make four adjustments to net zero public spending. First, the whole economy costs

than three. Second, they are increased by 25 per cent in aggregate during the period
2030 to 2050 to reflect cost savings associated with large-scale deployment being
realised more slowly, as well as pressures on supply chains being greater from a faster
transition. Third, further additional costs are included to reflect higher cost removals in
respect of higher residual emissions from sectors where it proves impossible to
decarbonise fully by 2050, such as domestic heating. Finally, operating savings are
concentrated into a period of two decades in line with deployment, but unlike for costs,
we do not vary the amounts saved. All told this raises the whole economy investment
by 21 per cent relative to the balanced pathway. We assume that half of the additional

84
The White House, Executive Order on Tackling the Climate Crisis at Home and Abroad, 27 January 2021.
85
Statement by H.E. Xi Jinping President of the People's Republic of China at
the General Debate of the 75th Session of The United Nations General Assembly, 22 September 2020.
86
European Council, European Council conclusions, 12 December 2019.
87
Financial Times, , 27 May 2021.
88
Burke, J., Sato, M., Taylor, C. and Li, F., What does an EU Carbon Border Adjustment Mechanism mean for the UK?, April 2021.

Fiscal risks report 142

 Climate change

whole economy cost is borne by the public sector, almost doubling net zero public
spending over the full period.

Receipts from emissions-intensive activities only begin to fall rapidly after 2031-32
reflecting the slower initial progress. enario in
which the transition to electric vehicles takes place more slowly, so that receipts from
fuel duty and vehicle excise duty fall a little less quickly than in the balanced pathway.

Carbon tax revenues are lower across the whole period due to full emissions coverage
at higher tax rates not coming into effect until 2031-32, though the tax rate quickly

emissions untaxed relative to that scenario. On an annual basis, additional revenues

move above the early action scenario from 2034-35 onwards thanks to the higher tax
rate and the higher path for gross emissions, with the latter falling more slowly than in
the early action scenario and remaining higher in 2050-51.

3.125 The debt-to-GDP ratio in the late action scenario is fractionally lower than the early action
scenario until the mid-2020s thanks to the slower pace at which emissions-related receipts
are lost. It then moves a little above the early action scenario in the second half of the
2020s because of the lack of additional carbon tax revenues. But the material differences
begin in the 2030s as a result of the economic disruption caused by the disorderly
imposition of more stringent policies, the higher cost of the transition that is borne by public
spending, and the lower long-term path for GDP. These combine to raise public spending to
1.2 per cent of GDP above the early action scenario on average in the two decades to
2050-51. This higher primary spending is partly offset by additional carbon tax revenues,
leaving borrowing between 0.9 and 1.8 per cent of GDP higher than the early action
scenario between 2031-32 and 2050-51. This leaves debt 23 per cent of GDP higher in
2050-51, roughly doubling the fiscal cost of bringing emissions down to net zero.

143 Fiscal risks report

Climate change

Chart 3.29: Early action versus late action scenarios: fiscal aggregates
54 54
Non-interest receipts Non-interest spending
52 52
Late action scenario
50 50
Early action scenario
48 48
Outturn
Per cent of GDP

Per cent of GDP

46 46
44 44
42 42
40 40
38 38
36 36
34 34
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
20 145
Borrowing Debt
135
125
15
115
105
Per cent of GDP
Per cent of GDP

10 95
85

5 75
65
55
0 45
35
-5 25
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
Source: ONS, OBR

The potential consequences for productivity

3.126 Over the long term, productivity growth is the key determinant of fiscal sustainability since it
underpins growth in all the major tax bases, which in turn provide the resources to meet
demands for public spending. Even small differences in productivity growth cumulated over
extended periods can have material consequences for the public finances. The effects of
global warming and decarbonisation on productivity are uncertain, as discussed in
paragraphs 3.121 to 3.122 above. There are examples of low-carbon technologies whose
costs have fallen so far and so quickly that they are now cheaper than the fossil-fuel
technologies they replace (as with solar and wind power, for example)89, and others where
that is expected to be the case in the relatively near future (as with electric cars). So an
optimistic scenario in which decarbonisation actually boosts productivity is not impossible.90
But equally, the costs of such a major structural change over a sustained period could easily

89
Our World in Data, Why did renewables become so cheap so fast? And what can we do to use this global opportunity for green
growth?, 2020.
90
Several studies have considered channels along which the transition to greener technologies could raise productivity. For example,
Mealy, P. and Teytelboym, A., Economic complexity and the green economy, 2021, uses network analysis to demonstrate that it is easier
for countries to become competitive in new green products that require similar production capabilities and know-how to existing sectors,
while Martin, R., Unsworth, S., Valero, A. and Verhoeven, D., Innovation for a strong and sustainable recovery, 2020, compare broad
categories of green technologies and find that the UK is relatively specialised in ocean and wind energy.

Fiscal risks report 144

 Climate change

be greater than assumed for example, if policy needs to change course unexpectedly or if
a technology that has been heavily invested in proves unsuccessful.

3.127
transition risks over the next 30 years, with real GDP growth over the coming decade a little
over 0.1 per cent points a year lower than in the baseline. To test the sensitivity of our fiscal
scenario results to different productivity paths, we therefore vary real GDP growth by 0.1
percentage points a year to either side of the early action scenario giving 3 per cent
differences in the level of GDP in 2050-51. This means real GDP settles at 1.6 per cent
above
below it in the downside variant. The latter is in line with the late action scenario in 2050,
but involves none of the intervening macroeconomic disruption, nor any of the additional
direct fiscal consequences included in that scenario.

3.128 In the high productivity variant, the larger economy, and the lower spending as a share of
GDP that results, lowers debt by 11 per cent of GDP relative to the early action scenario by
2050-51. This difference includes the modest debt interest savings associated with lower
primary spending. The results are broadly symmetrical in the low productivity variant.

Chart 3.30: Alternative productivity variants: real GDP and the debt-to-GDP ratio
2 135
Real GDP Debt
125
Percentage deviation from baseline

1
115

0 105
95
Per cent of GDP

-1 85
75
-2
65
-3
Low productivity variant
55
High productivity variant
45
-4 Early action scenario
35
Outturn
-5 25
2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
Source: ONS, OBR

High versus low public sector share of net zero investment

3.129 As discussed earlier in the chapter, there is considerable uncertainty over the whole
economy investment costs of the transition and the degree to which those costs will be borne
by the public sector. Even holding the whole economy costs constant at those implied by the

2050-51 being 5.9 per cent of GDP above the early policy action scenario; while the low
variant results in debt being 5.2 per cent of GDP below that scenario (Chart 3.31).

3.130 This simple sensitivity analysis does not incorporate any feedback from the higher or lower
paths of public investment to GDP growth or the cost of low-carbon technologies. Given the
various market failures and distortions that hold back private investment in decarbonisation,

145 Fiscal risks report

Climate change

it could be the case that early public investment yields future fiscal benefits from more rapid
deployment, lower costs and higher productivity than would otherwise have been achieved.
But, as with early action in general, it could also be the case that investment is wasted on
the wrong technologies or on those whose costs fall rapidly due to global developments.

Chart 3.31: Net zero public spending variants: spending and debt-to-GDP ratios
54 125
Non-interest spending Debt
52 115

50 105
48 95

Per cent of GDP

Per cent of GDP

46 85
44 75
42 65
Low public spending share
40 55
High public spending share
38 45
Early action scenario
36 35
Outturn
34 25
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
Source: ONS, OBR

Potential for offsetting fiscal policy adjustments

3.131 Two important assumptions in our early action scenario relate to future policy settings where
there is currently no clear statement of long-term government policy namely, those
relating to public investment and to taxes on motoring. In essence, our scenario reflects both
in their most costly form all net zero public investment is additional to the baseline and all
loss of fuel duty and vehicle excise duty is reflected in higher borrowing. So, another
sensitivity worth illustrating is the effect of potential offsetting policy adjustments. We can do
this by instead defining long-term policy assumptions in their least costly way, with all net
zero public investment allocated from within existing budgets and with the tax burden on
motoring maintained as existing taxes fall away. (The former might be more plausible at the
current juncture than it would have been in recent decades since the Government plans to
sustain public investment at multi-decade highs in the coming years.)

Public investment
3.132 Our baseline assumes that public sector net investment (PSNI) will continue throughout the
period at the 2.7 per cent of GDP it reaches in 2025-26. As net zero public investment
averages 0.4 per cent of GDP across the period and peaks at 0.5 per cent of GDP in 2028-
29, it need not all be additional to existing totals. Indeed, if, at the other extreme, all net
zero investment were allocated from within the baseline it would make up only 13 per cent
of PSNI on average over the period, peaking at 17 per cent in 2028-29.

3.133 If net zero public investment were all allocated from within the baseline instead of being
additional, PSNI would remain at 2.8 per cent of GDP throughout the period (a little higher
than the baseline due to the indirect effects from a smaller economy) rather than averaging
3.1 per cent of GDP. As a result, PSND would be 8.4 per cent of GDP lower by 2050-51

Fiscal risks report 146

 Climate change

than in our early action scenario, including the effects of lower debt interest spending (Chart
3.32). This would reduce the increase in debt at that point by around two-fifths.

Chart 3.32: Public investment additionality sensitivities

4.0 125
Public sector net investment Debt
115
3.5
105
3.0
95

Per cent of GDP

Per cent of GDP

2.5 85
2.0 75

1.5 65
Net zero investment from existing totals 55
1.0
Early action scenario 45
0.5
35
Outturn
0.0 25
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
Source: ONS, OBR

Taxes on motoring
3.134 Our scenarios assume that receipts lost through the decarbonisation of motoring lead to
progressively higher borrowing and debt. But since this is a predictable tax cut for motoring,
the Government could instead choose to levy a different tax to maintain revenues from

sixth Carbon Budget advice noted that this loss of tax revenues would have the side effect of
increasing congestion, and that to address both the fiscal cost and the greater congestion
some form of road pricing is likely to be necessary - could
apply to all vehicle types and be set at a level to fill the gap left by fuel duty 91 Indeed, the
Ten Point Plan for a Green Industrial Revolution we will need to
ensure that the tax system encourages the uptake of [electric vehicles] and that revenue from
motoring taxes keeps pace with this change 92

3.135 If lost receipts from fuel duty and VED were replaced by an equivalent yielding levy on
motoring (of whatever form), receipts in 2050-51 would be 1.5 per cent of GDP higher and
PSND would be 24 per cent of GDP lower than in the early action scenario (Chart 3.33).
This would more than offset the net impact of the transition on debt at that point, leaving the
debt-to-GDP ratio lower than the baseline. This illustrates how additional carbon tax
revenues in the early action scenario are sufficient to pay for the transition and its economic
consequences, so long as the tax burden on motoring is maintained.

91
See Box 6.5, Climate Change Committee, et Zero, December 2020.
92
The Ten Point Plan for a Green Industrial
Revolution, November 2020.

147 Fiscal risks report

Climate change

Chart 3.33: Motoring tax sensitivities

41 125
Non-interest receipts Debt
115
40
105
39 95
Per cent of GDP

Per cent of GDP

85
38
75
37
65

36 55
Motoring tax revenues maintained
45
35 Early action scenario
35
Outturn
34 25
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
Source: ONS, OBR

Combined sensitivities
3.136 Combining these two alternative formulations of our long-term policy assumptions (such
that net zero investment is allocated from within existing totals and fuel duty and VED
replaced by an alternative tax on motoring), PSND would be 32 per cent of GDP lower in
2050-51 than in the early action scenario (Chart 3.34). This would be 12 per cent of GDP
lower than the baseline, illustrating the extent to which additional carbon tax revenues and
use of the existing public investment envelope to fund net zero investment could limit the
s emissions to net zero.

Chart 3.34: Long-term policy assumption sensitivities: debt-to-GDP ratio

130

120

110

100

90
Per cent of GDP

80

70
Outturn
60
Early action scenario
50
Net zero investment from existing totals
40
Motoring tax revenues maintained
30
Net zero investment and motoring tax combined
20
2000-01 2005-06 2010-11 2015-16 2020-21 2025-26 2030-31 2035-36 2040-41 2045-46 2050-51
Source: ONS, OBR

Fiscal risks report 148

 Climate change

Summary of alternative scenarios and sensitivities

3.137 The fiscal implications of different paths to net zero vary widely. The impact on public sector
net debt in 2050-51 ranges from a decrease of 12 per cent of GDP in the most fiscally
favourable scenario (in which early action is combined with funding net zero investment
from existing totals and maintaining the tax burden on motoring), up to an increase of 43
per cent of GDP in the most fiscally costly one (in the late action scenario that combines
both higher investment costs and economic disruption due to policy measures being
introduced abruptly) (Chart 3.35). In virtually all variants, the largest moving parts relate to
revenues the receipts lost from decarbonisation and those gained from taxing all
emissions and at higher rates. The direct cost of public spending on the transition is modest
by comparison in all but the late action scenario and the high spending variant. But in terms
of the largest source of difference between the scenarios relative to the early action
reference point, it is the path of GDP and the indirect fiscal consequences from maintaining
public services and non-climate-related investment in real terms that is most important. This
illustrates once more the vital importance of productivity growth for fiscal sustainability.

Chart 3.35: Contributions to PSND scenario differences from baseline

70
Denominator effect Net zero receipts losses
60
Carbon tax revenues Net zero public spending
50
Indirect effects Debt interest spending
40
Overall effect
Per cent of GDP

30

20

10

-10

-20

-30
Investment Motoring tax High Net zero Low spending Early action High Low Late action
included and revenues productivity investment variant scenario spending productivity scenario
motoring maintained variant from existing variant variant
maintained totals
Source: OBR

3.138 Our fiscal scenario results and their relative positions reflect the assumptions that underpin
them, which vary in their sophistication. Refining them will form part of our future work on
climate change, including as the Government sets out the further policy measures that will
be necessary to meet its net zero target. Table 3.3 records the key assumptions that
underpin each scenario. The components of each are available on our website to allow
users to vary them or to combine elements from different scenarios to prepare their own.

149 Fiscal risks report

Climate change

Table 3.3: Summary of climate-related scenario assumptions

Real GDP CCC
OBR public OBR carbon-
(per cent scenario for
spending related OBR carbon
deviation whole
share revenue loss tax variant
from economy
variant variant
baseline) costs1
Early action scenario -1.4 Balanced Central Central Early
Late action scenario -4.6 Late Late Headwinds Late
High productivity variant 1.6 Balanced Central Central Early
Low productivity variant -4.4 Balanced Central Central Early
High spending variant -1.4 Balanced High Central Early
Low spending variant -1.4 Balanced Low Central Early
Net zero investment from existing totals -1.4 Balanced Central Central Early
Motoring tax revenues maintained -1.4 Balanced Central Central Early
Investment included and motoring maintained -1.4 Balanced Central Central Early
1'
Balanced' is the balanced net zero pathway

Conclusions
3.139 Over the past year, we have seen how a pandemic and the policy measures necessary to
bring it under control have changed our daily lives. We all hope that these changes will
prove to have been temporary. Unmitigated global warming has the capacity to deliver
catastrophic changes to lives and livelihoods and would be essentially irreversible.

3.140 Twenty-four years on from the Kyoto climate agreements, global emissions have yet to peak
and global temperatures have been rising unusually quickly. Against this backdrop, the
2015 Paris agreement, and the national targets flowing from it, are designed to limit further
global warming and to mitigate against the worst of its effects. The UK is among 131
countries that now have in place or are considering net zero emissions targets. But even on
more optimistic paths that assume policies come on stream for those targets to be met,
some further warming, and some associated economic costs, can be expected as well as

Between now and 2050, the fiscal costs of reducing net emissions to zero in the UK
could be significant but not exceptional. The CCC puts the cumulative 30-year
investment cost for the whole economy, plus the operating costs of removals, at £1.4
trillion in real terms, with our central variant assuming that the Government picks up
around a quarter of that cost. When combined with savings from more energy-efficient
buildings and vehicles, the net cost to the state is £344 billion in real terms. But spread
across three decades, this represents an average of just 0.4 per cent of GDP in
additional public spending each year. Factoring in the costs of lost fuel duty and other
emissions-related revenues, and the fiscal impact of a modestly smaller economy,
partly offset by the yield from taxing carbon more heavily, the fiscal impact of
achieving net zero would add 21 per cent of GDP to public sector net debt in 2050-51
(£469 23 per cent of

Fiscal risks report 150

 Climate change

GDP (£520 019-20 and 2021-22 that we

expect as a result of the pandemic.

By international standards, the UK has made good progress in reducing emissions, but
there are greater challenges ahead. As of 2019, UK emissions were down 44 per cent
relative to 1990. In particular, the source of power generation with the highest
emissions coal has disappeared from the energy mix thanks to concerted policy
-fired power
generation very heavily. Decarbonising other sectors will present many technological
and delivery challenges. As regards technology, the challenge is perhaps greatest in
removals particularly the direct air capture variety that is not yet available at scale. As
regards delivery, the challenge is perhaps greatest in domestic heating thanks to the
need to upgrade insulation and replace gas boilers and other fossil-fuel heating
systems in more than 28 million homes. This accounts for a fifth of whole economy
investment costs and the limited success of subsidy schemes introduced over the past
decade suggests that cost is not the only challenge for policymakers to overcome.

The costs of failing to get climate change under control would be much larger than
those of bringing emissions down to net zero. Our stylised unmitigated warming
scenario shows debt spiralling up to around 290 per cent of GDP thanks to the cost of
adapting to an ever hotter climate and of more frequent and more costly economic
shocks (as the spillovers from increased conflict and mass migration are added to the
cost of more extreme weather events). Viewing the costs of achieving net zero in this
context, it is clear the net benefits of a successful global response would be huge.

The UK accounted for just 1 per cent of global emissions in 2019, whereas China
accounted for 24 per cent and the United States for 12 per cent. So, the fiscal risks

There could be significant fiscal benefits from transitioning to net zero sooner rather
than later, not least the additional revenues that would come from taxing all emissions
at higher rates. Our early action scenario assumes that additional carbon tax revenues
start in 2026-27 and the resulting revenues to 2050-51 are sufficient to cover the cost
of the public investment to get to net zero more than twice over. Early publicly led
action could also overcome the inertia that slows decarbonisation in some sectors. But
it would come with risks too such as backing the wrong technologies or paying more
for the right ones than would be the case if global developments push costs down.

In the longer term, the largest fiscal cost of achieving net zero is the loss of fuel duty
receipts. In effect, this is a large and predictable tax cut on motoring one that would,
all else equal, increase congestion on roads. Maintaining the tax burden on
motoring as the Government has suggested it will need to do would therefore
address both the fiscal and congestion risks from this aspect of the transition.

151 Fiscal risks report

Climate change

3.141 Finally, while there is no uncertainty around the fact that climate change is happening or
what drives it, there is considerable uncertainty around the precise path for global
temperatures and their economic consequences, and the trends and policies that will
influence the transition to net zero over the next three decades. Will policy settings in the UK
and globally evolve to match the emissions targets that are being set? How quickly will
technologies evolve and their costs fall? What will prove to be the right balance between
taxing carbon and pulling other policy levers in incentivising decarbonisation? How
effectively can the large-scale, multi-year processes involved in decarbonising millions of
buildings and other infrastructure be managed? As understanding of these and other issues
improves, the uncertainty around the fiscal risks from climate change should abate.

Fiscal risks report 152

4 Cost of public debt

Introduction
4.1 The stock of government debt is both the result of past fiscal risks crystallising and a source
of future fiscal risks. As noted in the introduction to this Fiscal risks report (FRR), UK public
sector net debt has almost quadrupled since the turn of the century, rising from 27 per cent
of GDP in 2000-01 to an expected 107 per cent of GDP by the end of 2021-22.1 This partly
reflects discretionary decisions on the part of previous governments to run a looser fiscal
policy in normal times, often justified as a means of increasing public investment, which has
risen from 0.3 per cent of GDP in 2000-01 to 2.7 per cent of GDP this year. However,
about two-thirds of that 80 per cent of GDP increase in debt occurred in the immediate
aftermath of two major economic shocks: the 2008 financial crisis and the 2020
coronavirus pandemic.2 The risks that this elevated stock of debt itself poses to the fiscal
outlook depends in part on the future path of interest rates and the speed with which any
change in interest rates is reflected in the public sector

4.2 Debt interest costs amounted to 4.1 per cent of total public spending in 2019-20, down
from 9.8 per cent in 1980-81.3 They fell to 2.1 per cent in 2020-21, as debt interest costs
fell and fiscal support measures pushed spending up dramatically, but we expect them to
remain at a historically low 3.0 per cent of spending in 2025-
effects have largely passed. Debt interest costs reflect the stock of debt in issue and the
interest paid thereon, which tends to vary with the maturity of each debt instrument. The
debt maturity structure also determines how quickly changes in market interest rates feed
through to debt interest costs. The Government also receives interest on its financial assets,
which is determined by similar factors. Over the past three decades net interest payments by
the Government have fallen sharply as a share of GDP, from 3.8 per cent in 1980-81 to
0.9 per cent in 2020-21, despite the debt-to-GDP ratio rising sharply from 40.4 per cent to
100.2 per cent (Chart 4.1). This reflects the downward drift in short and long-term interest
rates to historically low levels, both in absolute terms and relative to the growth rate of GDP,
a phenomenon common to many advanced economies in recent years.

4.3 Despite this decline in interest rates and costs over recent decades, there is considerable
uncertainty around their future path. And higher post-pandemic government debt,
combined with a shortening of its effective maturity as a by-product of quantitative easing,
more exposed to increases in government borrowing costs.
Were rates to return to levels that were more normal in the past, it would raise the cost of
servicing a given stock of debt and could in extreme circumstances push the debt-to-
GDP ratio onto an unsustainable path.

1
All figures for 2020-21 onwards used in this chapter are as forecast in our March 2021 Economic and fiscal outlook.
2
Increase in the debt-to-GDP ratio in 2008-09 and 2009-10 for the financial crisis and in 2020-21 and 2021-22 for the pandemic.
3
Measured as central government debt interest net of the Asset Purchase Facility as a percentage of total managed expenditure.

153 Fiscal risks report

Cost of public debt

Chart 4.1: Debt to GDP, the growth-corrected interest rate and net interest payments
WWI WWII Post-war decades Post-financial crisis Pandemic
300
Debt-to-GDP ratio
250

200
Per cent of GDP

150

100

50

0
30 1 11 21 31 41 51 61 71 81 91 101 111 121

20

10
Per cent

-10

Year-on-year change in nominal GDP (G)

-20
Effective nominal interest rate (R)
Growth-corrected interest rate (R-G)
-30
10 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 101 106 111 116 121 126 40

9 35
Net interest payments (left axis)
8
Net interest payments (right axis) 30
7
Per cent of revenue
Per cent of GDP

6 25

5 20
4 15
3
10
2
1 5

0 0
1900-01 1910-11 1920-21 1930-31 1940-41 1950-51 1960-61 1970-71 1980-81 1990-91 2000-01 2010-11 2020-21
Notes: For all charts, 2020-21 onwards shows our March 2021 forecast. The debt-to-GDP ratio is PSND. The effective nominal
interest rate is net interest payments divided by PSND. Net interest payments are interest payments less interest received. R evenue
is net of interest receipts.
Source: Bank of England, ONS, OBR

Fiscal risks report 154

 Cost of public debt

4.4 Indeed, having fallen sharply at the onset of the pandemic, long-term interest rates have
subsequently begun to rise as the successful rollout of effective vaccines has raised the
prospect of a rapid re-normalisation of economic activity. Since the start of this year, UK
Government 10-year bond yields have risen by around 0.6 percentage points. In part
reflecting the particularly expansionary monetary and fiscal stance in the US, market
participants have also increasingly focused on the risk of a reignition of inflation, although
to date the rise in inflation expectations appears to have been much less here than in the
US.4

4.5 Against this backdrop, in this chapter we cover:

the risks from high government debt and the dynamics of its evolution, including the
-

drivers of the recent fall in real interest rates and implications for future trends in
government bond yields;

the fiscal implications of a rise in global real interest rates;

the fiscal implications of higher UK inflation; and

the fiscal implications of a loss of investor confidence in UK sovereign debt.

4.6 In the later parts of this chapter, we present several scenarios illustrating the consequences
of assuming different paths for borrowing costs, inflation and GDP growth for the public
finances. The purpose of these scenarios is to expose the mechanisms at work and the
approximate quantitative magnitudes involved, rather than to provide precise modelling of
specific events. In addition, although the risks are in most cases two-sided, we focus mainly
on scenarios that lead to a deterioration in the public finances, as these would be likely to
present greater challenges for the Government. The loss of investor confidence scenario
illustrates the crystallisation of an extreme tail risk, in line with the focus of this report on
such events.

Government debt risks and dynamics

4.7 Higher levels of debt expose governments to greater fiscal risks. Financing spending by
borrowing rather than taxes enables governments to spread the burden of responding to
shocks such as wars, financial crises and pandemics over time and thus share it with
future generations of taxpayers. But the willingness of investors to hold a
will depend on their confidence in its ability and willingness to undertake the fiscal actions
necessary to meet its debt obligations as they fall due. Other things equal, the fiscal costs of
meeting those obligations (and risk the Government might not be able to meet them) will be
greater the higher the debt stock relative to the size of the economy. The willingness of

4
Vlieghe, G., What are government bond yields telling us about the economic outlook?, Bank of England Speech 27 May 2021.

155 Fiscal risks report

Cost of public debt

corporate bonds,
equities or real estate. So there will be a limit on the capacity of a government to borrow
and high debt may constrain that willingness and ability to undertake
desirable fiscal actions for fear of reaching this limit (see Box 1.1 fiscal
space ).

4.8 Well before that upper limit on borrowing is reached, however, investors are likely to
become more concerned about the risk that the Government will fail to meet its obligations,
through either outright default, artificial suppression of nominal interest rates via regulatory
, or reducing the real value of its debt obligations through
higher inflation (where the country has its own currency in which its debt is denominated).
The UK Government has never formally defaulted on its marketable debt, although there
5

Additionally unanticipated inflation did play a major role in re

post-war debt, especially during the 1970s.

4.9 Investor concerns about outright default or future erosion of the real return on government
debt may lead to a higher cost of borrowing.6 In extremis, where investors lose confidence
a government can face a total loss of
more common
to emerging markets and developing countries though not unknown in advanced
economies, governments that cannot access sufficient emergency financing from bilateral or
multilateral sources would need to cut spending and/or raise taxes sharply if they are to
meet their debt obligations.

4.10 debt, the more sensitive the public finances become to

increases in interest rates. But the maturity structure of that debt is a crucial determinant of
the speed at which higher interest rates feed through into higher interest costs, with shorter
average maturities leading to faster pass-through. Debt levels and maturities can also
interact, as heavily indebted governments may seek to shorten average maturities in order
to take advantage of the typically upward sloping yield curve for government debt. In
extreme circumstances, this can fuel a spiral of shortening maturities and growing sensitivity
to further interest rate rises, eventually culminating in loss of market access or default.

4.11 Short of these cataclysmic events, higher government debt also appears to be associated
with slower GDP growth over the long run. Most of the empirical evidence finds such an
inverse relationship between government debt and GDP growth, although the quantitative
magnitudes vary.7 This could arise because the competition for funds drives up the cost of

higher expected future taxes. The causality could also run in the other direction, with lower
GDP growth leading countries to accumulate more debt.

5
Ellison, M., and Scott, A., Managing the UK National Debt 1694-2018, American Economic Journal: Macroeconomics, 2020.
6
See discussion of the impact of haircuts in sovereign restructurings on subsequent bond spreads in Cruces, J., and Trebesch, C.,
Sovereign Defaults: The Price of Haircuts, American Economic Journal: Macroeconomics, 2013.
7
For example, see de Rugy, V., and Salmon, J., Debt and Growth: A Decade of Studies, Mercatus Center Policy Brief, April 2020.

Fiscal risks report 156

 Cost of public debt

4.12 So high debt is potentially a cause of concern, especially since the UK government debt-to-
GDP has risen from the eighteenth highest amongst advanced economies in 2001 to the
eighth highest in 2020.8 Despite that, and as already noted, the cost of debt service has
declined to historic lows as a result of the decline in the yields on government debt. To
illustrate the dynamics involved, we start by noting that, in addition to the initial stock of
debt, the identity describing the evolution of the debt-to-GDP ratio reflects three factors:9

First, the size of the primary balance the difference between government spending
on everything except debt interest, and tax revenues and other receipts net of interest
received by the Government.

Second, the -corrected interest rate - the difference between the

nominal interest rate paid on government debt (R), which pushes up the debt-to-GDP
ratio, and the growth rate of nominal GDP (G), which reduces it. R is measured as net
interest payments divided by the face value of the stock of government debt (i.e. the
effective interest rate on that debt). R is thus the average interest rate that the
Government pays on its stock of debt each period, rather than the interest rate
prevailing in the market (which would represent the marginal cost of any new debt
issued by the Government).

Third, stock-flow adjustments changes in debt unaccounted for by the primary

balance or debt interest. These are usually the consequence of the net acquisition of
financial assets, or of timing, classification and valuation effects.10

4.13
measured at face value Box 4.1 discusses the measurement of government debt in more
detail.

Box 4.1: Face and market value of debt securities in official statistics
The National Accounts framework recognises three possible ways of valuing government bonds:

market value, which represents the amount the Government would have to pay to buy
back the stock today;

face (or redemption) value, which is the amount that the Government has promised to
pay to bond holders when the bonds mature; and

nominal value, which is the original exchange value adjusted for any subsequent
payments or accrued interest.

8
Based on general government net debt to GDP ratios in 32 advanced economies for which the IMF provides data.
9
This can be expressed as: 𝑑 − 𝑑 = 𝑝 + 𝑠 + [(𝑅 − 𝐺 )/(1 + 𝐺 )]𝑑 . The change in the debt-to-GDP ratio (𝑑 − 𝑑 ) is equal to the
primary deficit (𝑝 ), plus any stock-flow adjustments (𝑠 ), plus the impact of any difference between the effective nominal interest rate (𝑅 )
on the debt stock and nominal GDP growth (𝐺 ).
10
Examples of the former include loans issued to the private sector (the financing of which adds to debt but not the accrued deficit since
they are matched by an asset). Examples of the latter include: the lag between tax liabilities being incurred and paid (which vary from tax
to tax); changes to the public sector boundary that bring liabilities into or out of scope of measured public debt (as with reclassifications of
housing associations in recent years); and currency movements that change the sterling value of the foreign exchange reserves.

157 Fiscal risks report

Cost of public debt

Face or market value are most commonly used. In the past decade these two measures have
diverged sharply, with the gap for gilts held by the private sector reaching about 15 per cent of
GDP in 2020-21. Some of this has been caused by the increased stock of debt (since a given
proportionate difference has increased when expressed as a share of GDP), but mainly it has
been driven by declining yields and therefore increasing prices for government debt. This has
increased the market value of the existing portfolio, while also affecting the price received for new
debt issuance (particularly for index-linked gilts). The DMO prefers to issue new gilts with coupon
payments cl
and market value at the time of issuance will be similar. However, as real interest rates are now
negative, index-linked gilts would need to be issued with negative coupons to achieve a par
price. This is not practicable and so prices for index-linked gilts are at a significant premium to par.

Chart A: Face versus market value of gilts held by the private sector
100
Face value
90
Market value
80

70

60
Per cent of GDP

50

40

30

20

10

0
1999-00 2001-02 2003-04 2005-06 2007-08 2009-10 2011-12 2013-14 2015-16 2017-18 2019-20
Source: ONS, OBR

The accounting identity that describes the evolution of the debt stock can be written in terms of
either of these two debt valuations. The official measure of public sector net debt that we are
charged with monitoring is measured at face value, so we use that definition in the analysis in
this chapter. An alternative approach is to write the accounting identity in terms of market value,
in which case the return on bonds includes not only coupon payments, but also capital gains or
losses. This would be more suitable for some other purposes, such as evaluating the value for
money of past debt issuance choices (which is beyond our remit). Papers by Hall and Sargent,
and Scott and Ellison, provide a fuller discussion of the connection between the two approaches,
as well as time series for US and UK government debt under the market value approach.a

Error! Reference source not found. Hall, G., and Sargent, T.J., Interest rate risk and other determinants of post WWII U.S. government
debt/GDP dynamics, American Economic Journal: Macroeconomics, 2011, and Ellison, M. and Scott, A., Managing the UK National
Debt 1694-2018, American Economic Journal: Macroeconomics, 2020.

Fiscal risks report 158

 Cost of public debt

4.14 The growth-corrected interest rate plays a particularly important role in debt dynamics and
the analysis of debt sustainability. When the interest rate exceeds the rate of growth, extra
debt incurred as a result of a temporary rise in the primary deficit must ultimately be paid
for by higher taxes (or lower spending) in the future, otherwise the debt-to-GDP ratio will
rise indefinitely. But, as recently pointed out by Blanchard, if the interest rate exceeds the
rate of growth, the Government can pay for both the interest and principal by issuing more
debt without triggering an upward spiral in the ratio of debt to GDP.11 Blanchard also notes
that the growth-corrected interest rate on US government debt has often been negative,
including today and the recent past. Were that to continue to be the case, then the fiscal
costs of extra debt would be negligible (though not necessarily the economic costs, as
Blanchard explains).

4.15 The growth-corrected interest rate paid on UK gilts has also been negative for much of the
post-war period, including for most of the past decade, where the decline in yields since the
1990s has been greater than the fall in nominal GDP growth over that period (as shown in
the middle panel of Chart 4.1 earlier in the chapter). This has led to historically low debt
servicing costs for the Government, despite the debt-to-GDP ratio reaching its highest level
since the early 1960s. Between 1997-98 and 2020-21, the effective interest rate on
government debt has fallen from 7.2 per cent to 1.1 per cent.

Global interest rates

4.16 Whether the UK is likely to see continued low borrowing costs in future or a reversal
depends in part on the factors driving down interest rates in the past. It is worth noting at
the outset that the fall in interest rates on government debt is not just a UK but a global
phenomenon, with long-term nominal government bond yields in all the major advanced
economies drifting steadily downwards from an average of 9.0 per cent in 1990 to just 0.2
per cent in 2020 (Chart 4.2, left panel).12 Other countries have also experienced falling
nominal GDP growth, though not to the same extent as the fall in bond yields (right panel).
The commonality in these movements strongly suggests that global factors, or domestic
ones that are common across countries, have been at work.

11
Blanchard, O., Public debt and low interest rates, American Economic Review, 2019.
12
Unweighted average of nominal 10-year bond yields in Germany, USA, UK and Japan.

159 Fiscal risks report

Cost of public debt

Chart 4.2: 10-year nominal government bond yields and nominal GDP growth
16 16
10-year nominal government bond yields Nominal GDP growth
14 14

Percentage change on a year earlier

12 12
10 10
8 8
6 6
Per cent

4 4
2 2
0 0
UK UK
-2 -2
Germany Germany
-4 -4
USA USA
-6 -6
Japan Japan
-8 -8
1985 1990 1995 2000 2005 2010 2015 2020 1985 1990 1995 2000 2005 2010 2015 2020
Source: OECD Source: IMF

4.17 We can split nominal bond yields into the real bond yield plus inflation expectations. The left
panel of Chart 4.3 shows nominal and real yields for high-quality government bonds (as
represented by a weighted average across the US, eurozone, Japan and the UK), with the
difference between them an indicator of expected inflation.13 This suggests that although
declining inflation expectations partly explain the fall in nominal rates from 1985 to the
mid-1990s, the decline since then appears to have been primarily a real, rather than
nominal, phenomenon.

4.18 At the same time, the return on capital (proxied by the yield on global equities in the right
panel of Chart 4.314) does not appear to have fallen in the same way as the real yields on
high-quality government bonds. While they fell together during the 1980s and 1990s, the
two have diverged since the turn of the century with equity yields rising (albeit with
significant volatility around the financial crisis and the pandemic), whereas government
bond yields have continued to fall.

13
Where data are available, the nominal yield is on 10-year government bonds and the real yield is on 10-year inflation-linked
government bonds for the US, Eurozone, Japan and UK. Where data on index-linked bonds are not available, the real yield is estimated
using the relationship between variables that are available over a longer period (10-year nominal government bond yields and current
and lagged inflation). To calculate the composite series, yields from each country are weighted by nominal GDP at current exchange
rates. The difference between the two is not a perfect measure of expected inflation because illiquidity in the conventional and index-linked
gilt markets could distort the
14
The earnings yield is calculated as 𝜋 /𝑉 , where 𝜋 is the earnings of all quoted companies gross of net interest payments and
corporate tax in country i in the year to time t and 𝑉 is the total market value of those companies including equity plus net debt.

Fiscal risks report 160

 Cost of public debt

Chart 4.3: Selected high-quality government bond yields and the return on capital
16 16
Real and nominal bond yields Return on capital and real bond yields
14 14
Real 10-year bond yield Real 10-year bond yield
12 12
Nominal 10-year bond yield Earnings yield on global equities
10 10

Per cent
8
Per cent

8
6 6
4 4
2 2
0 0
-2 -2
1985 1990 1995 2000 2005 2010 2015 2020 1985 1990 1995 2000 2005 2010 2015 2020
Note: The nominal and real 10-year bond yields are for the US, Eurozone, Japan and UK weighted by nominal GDP at current
exchange rates. Countries are included in years where data is available.
Source: Daly, K., A Secular Increase in the Equity Risk Premium, International Finance, 2016 (Data extended to Q2 2021), OECD, OBR

4.19 So any explanation of the recent decline in government bond yields, and prospects for
future reversal in this trend, needs to be compatible with three observations:

First, that the fall in the nominal yields on government debt since the mid-1980s has
been a persistent, global trend.

Second, while falling inflation expectations following the stabilisation of inflation in

part explains the fall in yields in the first part of the period, the declines since the mid-
1990s appear to be largely a real phenomenon.

And third, returns on riskier assets have not fallen in the same way, and indeed have
on average risen, over the past two decades.

Box 4.2 describes a simple analytical framework to help understand this set of facts and
how various factors are likely to affect the yields on high-quality bonds and on riskier assets
such as capital.

Box 4.2: A simple analytical model of the capital market

To help explain the forces that have generated the movements in yields observed in recent
decades, it is helpful first to construct a stylised model of the global capital market, shown in
Figure A.a There are two assets: capital (K), whose real return is uncertain; and government
bonds (B), whose real return is certain. In practice, of course, the return on government bonds
will be uncertain too because of the possibility of default, either de jure or de facto through
inflation, but for now it is helpful to put this to one side.
The overall supply of savings, or equivalently the demand for assets, (SS) that has to be held in
one of these two forms derives from the need for households to provide for their retirement and
to smooth consumption. It therefore depends on income, together with factors such as expected

161 Fiscal risks report

Cost of public debt

longevity and retirement ages and whether there are unfunded pension schemes in place that
affect the need for savings. It will also be affected by the expected real return on those savings,
rp, which is an appropriately weighted average of the expected real return on capital (rK) and the
real return on bonds (rB). The supply of assets then derives from: the demand for capital by
businesses for investment (II), which, in turn, depends on factors such as expected productivity
and the required return on those funds, rK; and the supply of bonds, which we take as
exogenous. A possible equilibrium outcome is depicted in the left-hand panel of Figure A, which
is shown assuming that savings increase as the rate of return on savings increases i.e. the SS
curves slopes upwards (note that the analysis would be the same if SS sloped downwards so long
as it is steeper than II).
In order to see how the returns on risky and safe assets are related and move together, it is
helpful to look at the right-hand panel of Figure A. The downward-sloping line AA shows the
combinations of rK and rB that are consistent with overall asset market equilibrium (i.e. where the
total supply of savings is equal to the total demand for them), other things equal. It slopes down
because a lower required return on capital raises the demand for capital by businesses but also
lowers the overall supply of savings. To bring forth the necessary extra savings, bonds would
then need to offer a suitably higher return so that the overall expected return on the portfolio is
sufficiently high to return the market to equilibrium.
We then need to supplement this with another, upward-sloping, relationship (PP) that shows the
combinations of rK and rB that are consistent with portfolio equilibrium (i.e. that ensure the

the respective supplies of each). In simple finance models, the spread of rK over rB, also known
as the equity risk premium, depends just on the statistical properties of the returns on capital and
the risk appetite of investors. But in arguably more realistic settings with incomplete markets and
financial frictions, a greater range of factors may become relevant. In particular, government
bonds may offer not only safety but also liquidity services for instance, banks and other
financial institutions can usually offer high-quality government debt as collateral for borrowing
short-term funds from the central bank or other financial intermediaries. In such cases, the
premium may also be affected by asset supplies; in particular, if bonds are already very plentiful
the value of the extra liquidity services provided by additional issuance will be quite low.b
We can use this diagram to identify the sort of factors that are likely to have driven yields in
recent years. In the 1990s, the yields on bonds and on capital fell together, roughly one-for-one.
That is consistent with factors shifting the asset market equilibrium schedule AA inwards, so that
rK and rB move along the portfolio equilibrium schedule PP in a south-westerly direction.
hypothesis, which focuses on a chronic tendency of savings to
exceed investment, produces just such an outcome.c
Since the early 2000s, however, it appears that the return on capital has been edging up at the
same time as bond yields have continued to decline, so that rK and rB have been moving in a
north-westerly direction. To explain this, one needs to invoke upward shifts in the portfolio
equilibrium schedule, PP, reflecting a shift in the demand and/or supply of assets in favour of
safe assets and away from risky assets (it is possible, of course, that the AA schedule has at the
same time continued to shift inwards).

Fiscal risks report 162

 Cost of public debt

Figure A: Determination of rates of return

Rate of Expected
return Capital market return on Equilibrium rates
SS capital PP
6

5
rK
4

Risk premium
3

1 AA
II
45°
0
Assets 1 rB Return on
bonds
K B

a
This is loosely based on the overlapping generations model discussed in Blanchard, O.J. Public Debt and Low Interest Rates,
American Economic Review, 2019.
.b For an analysis embodying these ideas, see Reis, R., The constraint on public debt when r<g but g<m, Centre for Economic Policy
Research Discussion Paper, March 2021.
.cRachel, L. and Summers, L.H., On falling neutral real rates, fiscal policy, and the risk of secular stagnation, Brookings Papers on
Economic Activity, 2019.

Explanations for the decline in real bond yields

4.20 Having established that the fall in interest rates on government debt over the past two
decades is a global, real phenomenon distinct from returns on other assets, this section
considers the potential explanations for the steady decline in real bond yields over time and,
by implication, the potential for any reversals in these trends that could put the public
finances under greater pressure. In particular, it considers the potential role of:

demographic trends affecting both aggregate savings and preferences between safe
and risky assets;

rising income inequality concentrating wealth in the hands of those with higher
propensities to save;

slower productivity growth, which reduces businesses desire to invest and raises
household demand for savings;

falling prices for physical capital reducing the funds necessary to purchase a given
quantity of equipment;

163 Fiscal risks report

Cost of public debt

increased risk awareness raising the demand for safe assets; and

lower supply of safe assets, in part reflecting changing investor perceptions of which
assets offer a reliable store of value.

4.21 Throughout this section, - to refer to assets where the

risk of default is seen as negligible, such as high-quality government bonds. In practice, no
asset can be completely safe and the real returns on nominal bonds will also be uncertain
because of inflation. Moreover, safety is not an intrinsic feature of an asset but depends
intimately on investors beliefs about the creditworthiness of the debtor. Investors may
collectively regard the bonds of a particular country as safe at one juncture, but later view
them as distinctly risky (this was the case, for instance, for euro area periphery government
debt during the financial crisis). We return to this issue in the final parts of the chapter.

Demographic trends
4.22
and will continue to do so for many years. Assuming average retirement ages do not rise
commensurately, that implies people will need to save more to fund more years spent in
retirement. Typically, the bulk of such savings is made by those in the later part of their
working lives, rather than by the young. Chart 4.4 shows the proportion of the global
population aged 40 to 64 (who are likely to be doing the bulk of the saving) compared to
the proportion of those aged 65 and over (who are more likely to be dissaving). The share
of middle-aged people has been growing steadily since the late 1980s, reflecting in part the
post- though those cohorts are now moving into retirement and will be
starting to dissave. This trend has been particularly marked in China, which is shown
separately,15 but wider global demographic trends are likely to have played at least as
important a role in boosting global savings over the past half-century. In addition, reduced
fertility has lowered the growth of the working-age population which reduces the investment
necessary to keep the labour force equipped. Taken together these demographic
developments should have raised desired savings relative to desired investment (pushing
AA down in Figure A of Box 4.2), lowering the real yields on both bonds and capital.

15
Bernanke, B. S. The global saving glut and the U.S. current account deficit, Sandridge Lecture, 2005.

Fiscal risks report 164

 Cost of public debt

Chart 4.4: Middle versus old-aged population globally and in China

50 50
World excluding China China
Middle-aged Projection Middle-aged Projection
40 40
Per cent of total population

Old-aged Old-aged
Difference Difference
30 30

20 20

10 10

0 0
1960 1980 2000 2020 2040 1960 1980 2000 2020 2040
Note: Middle-aged is defined here as 40-64 years, old-aged is over 65.
Source: World Bank, OBR

4.23 Demographic shifts could explain a general decline in yields and also appear to help
explain the relative stability of the return on capital and its rising spread over bond yields.
Many of those saving for retirement will be relatively risk-averse, while pension funds
offering defined benefits will often be required to hold matching assets in the form of bonds.
Moreover, those saving for retirement are frequently advised to steadily increase the
proportion of their wealth held in bonds as they grow older. So demographic developments
may have contributed to a shift in portfolio preferences towards bonds (an upward shift in
PP in Figure A of Box 4.2), which reduces the yield on bonds and raises the yield on capital.

4.24 Looking to the future, the proportion of old-aged people is set to rise faster than middle-
aged workers, reversing the trend since the 1990s (Chart 4.4). The implications of this for
the real rate on bonds are, however, unclear. Goodhart and Pradhan argue that this
demographic reversal will push up real interest rates as the dissaving of the elderly starts to
dominate the saving of the middle-aged.16 However, even if the overall rate of asset
accumulation falls, retirees will only run down their assets over many years (and indeed
rarely do so completely by the time they die), while also typically increasing the share held in
safe forms. So the upward pressure on bond yields from this source is likely to take many
years to materialise, though forward-looking investors may bring forward its effects.17

4.25 In summary, growing demand for safe assets from older workers may explain part of the
fall in yields on government bonds in recent years. But it is not clear that rising numbers of
retirees will lead to a rapid falling off in demand for government debt in the decades
ahead. Indeed, as life expectancy increases and people spend longer in retirement, this
could sustain demand for safer assets to fund their pensions. Therefore, the impact of
continued ageing of the global population on government bond yields is, at best, uncertain
and seems likely to take longer to materialise than some have suggested.

16
Goodhart, C., and Pradhan, M., The Great Demographic Reversal: Ageing Societies, Waning Inequality and an Inflation Revival, 2020.
17
See for example Auclert, A., Malmberg, H., Martenet, F., and Rognlie, M., Demographics, Wealth, and Global Imbalances in the
Twenty-First Century, mimeo Stanford University, 2020.

165 Fiscal risks report

Cost of public debt

Rising income inequality

4.26 A second factor that may have contributed to higher savings and the decline in bond yields
is increased inequality.18 Since the early 1980s, many countries have seen an increase in the
share of national income going to those with higher incomes. And since higher income
households tend to save more of their income than the less well off, this raises demand for
assets of all kinds (Chart 4.5). The reasons for the increase in income inequality are the
subject of considerable debate, but possible explanations include technological change

influence of organised labour, the growing importance of higher education for future
earnings, and changes in tax and benefit systems.19 However, increased income inequality
cannot easily account for the differential movement in the yields on bonds and capital since
2000, especially as one would expect wealthier households to have a greater appetite for
holding higher risk assets. And trends toward greater inequality have moderated over the
past decade at a time when government yields have continued to fall.

Chart 4.5: Income inequality and saving rates across the income distribution
50 60
Average saving rate by income quintile
Share of income accruing to the top 10

Income inequality
50
45
per cent of the distribution

40
40 30
Per cent

20
35
10
30 0
-10
25
UK US Germany -20
France Japan
20 -30
1961 1971 1981 1991 2001 2011 1 2 3 4 5
Income quintile
Note: Chart shows rolling five year averages Note: Results shown are for the US
Source: World Inequality Database Source: Dynan et al (2004)

4.27 The outlook for inequality is uncertain. The benefits of technological change and
automation may continue to accrue mainly to those on higher incomes. Against that, there
is some evidence that the pandemic may have made people less tolerant of inequality,20
which could manifest itself in more redistributive policy settings.

18
See for example Auclert, A., and Rognlie, M., Inequality and Aggregate Demand, mimeo, 2020.
19
See for example Dabla-Norris. E., Kochhar, K., Suphaphiphat, N., Ricka, F., and Tsounta, E., Causes and Consequences of Income
Inequality: A Global Perspective, IMF Staff Discussion Note, 2015.
20
Asaria, M., Costa-Font, J., and Cowell, F., COVID-19 has made us more averse to both income and health inequalities, 2021.

Fiscal risks report 166

 Cost of public debt

Slower productivity growth

4.28 Measured productivity growth has been weak in advanced economies since around the time
of the financial crisis, with the slowdown being particularly marked in the UK.21 Close
examination of the data, particularly for the US, suggests that the slowdown predated the
financial crisis, though it may have been aggravated by it.22 A likely consequence of weaker
productivity growth is a reduction in the propensity to invest by business. In addition, it will

Together, these act to raise the supply of savings relative to the demand for funds to invest
(so pushing AA down in Figure A of Box 4.2), lowering the yields on both bonds and capital.
Indeed, in some simple theoretical settings, yields and the (expected) growth rate should
move together one-for-one. This hypothesis fails, though, to provide an explanation for the
disparate movements in the yields on bonds and capital since the early 2000s.

4.29 Looking ahead, views on the outlook for productivity growth differ. At the pessimistic end,
Gordon argues that the past 250 years has been a period of unduly rapid growth based on
three major general-purpose technologies (the steam engine; electricity and the internal
combustion engine; and the digital revolution) that are now largely exhausted and that,
together with a plateauing in educational attainment, the pace of innovation is likely to be
permanently lower.23 At the other end of the spectrum, Brynjolfsson and McAfee argue that
the impact of the digital revolution is both underestimated in the official statistics and also
still has a long way to run.24 The central view embodied in our own EFOs and FSRs is for a
gradual revival in UK productivity growth, although not to the historically high rates seen
during the first part of the post-war period.25 A gradual productivity revival would result in
higher yields on bonds and capital (pushing AA up in Figure A of Box 4.2).

Falling relative price of capital goods

4.30 Another potential explanation for the decline in the demand for funds to invest is a falling
relative price for capital goods, reflecting the faster productivity growth in manufacturing
than in services and, more recently, in the information technology sector in particular.26 This
means any given investment project costs less, so the same investment volume can be
achieved by absorbing a smaller share of nominal GDP. A lower cost of capital should also
incentivise additional investments, so the net impact on desired investment depends on the
elasticity of the demand for capital with respect to its price, but empirical estimates suggest
the first effect dominates, so pushing down the net demand for funds to invest.27

4.31 Looking ahead, Eichengreen has argued that this downward trend in the relative price of
capital goods may slacken as technological developments allow faster improvements in

21
For a discussion on the reasons behind this fall see Goldin, I., Koutroumpis, P., Lafond, F., and Winkler, J., Re-evaluating the sources of
the recent productivity slowdown, 2021.
22
Antolin-Diaz, A., Drechsel, T., and Petrella, I., Tracking the Slowdown in Long-run GDP Growth, Review of Economics and Statistics, 2017.
23
Gordon, R. J., The Demise of US Economic Growth: Restatement, Rebuttal, and Reflections, 2014.
24
Brynjolfsson, E. and A. McAfee, Race Against the Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity,
and Irreversibly Transforming Employment and the Economy, Digital Frontier Press, 2011.
25
See Annex B of our March 2020 Economic and fiscal outlook for a fuller discussion of this assumption.
26
Karabarbounis, L., and Neiman, B., The Global Decline of the Labour Share, NBER working paper, 2014.
27
See discussion in Thwaites, G., Why are real interest rates so low? Secular stagnation and the relative price of investment goods, Bank of
England working paper, 2015.

167 Fiscal risks report

Cost of public debt

consumption goods and services.28 Consistent with this, the relative price of capital goods
appears to have recently stabilised, although difficulties in capturing quality improvements
in measures of the price of information technology goods remain.29 But as with slower
productivity growth, this hypothesis cannot explain the disparate movement in the yields on
bonds and capital since the early 2000s.

Increased preference for safety

4.32 A more convincing explanation for the divergence in the yields on government bonds and
capital since 2000 is a shift in investor preferences in favour of safer assets (i.e. an upward
shift in the PP schedule in Figure A of Box 4.2). In particular, increased awareness of the risk
of occasional large adverse
30
have lowered invest On top of that, tighter regulation since the
financial crisis has required banks and other financial institutions to increase their holdings
of safe assets. Coupled with changes in the relative supplies of risky and safe assets
discussed below, this could explain falling yields on government bonds relative to riskier
assets.

4.33 As far as the outlook goes, the coronavirus pandemic may have reinforced investor caution,
though that may be mitigated somewhat by the very substantial insurance provided through
the generous government support measures (see Chapter 2). One factor that may, however,
disturb this is the capital losses that will crystallise if bond yields do start to rise. While high-
quality government bonds such as US treasuries and UK gilts are most unlikely to default,
their market value could still fall and substantially so, given the current very low yields and
correspondingly high market values. Once investors start to experience significant capital
losses, there is a greater risk that they will take flight, pushing bond prices even lower and
yields even higher. The 1994 bond market crash provides a salutary reminder of what can
happen.31 So one should not altogether discount the possibility of a sharp correction to
bond yields (we return to this our final scenario from paragraph 4.94).

Lower supply of high-quality assets

4.34 The outstanding stock of government debt in advanced economies has increased sharply
over the past few decades, partly owing to the dual shocks of the financial crisis and the
pandemic. This should have acted to counteract the increased demand for safe assets
discussed above, limiting the downward pressure on bond yields.32 However, two
developments have offset that.

First, the financial crisis led to a narrowing in the class of assets viewed as safe in
particular, both AAA-rated securitised mortgage debt and eurozone periphery
government debt were shown to be far from safe.33

28
Eichengreen, B., Secular Stagnation: The Long View, NBER working paper, 2015.
29
See discussion in Rachel, L., and Smith, T., Secular drivers of the global real interest rate, Bank of England working paper, 2015.
30
Barro, R., Rare disasters and asset markets in the Twentieth Century, Quarterly Journal of Economics, 2006, and Daly, K, A secular
increase in the risk premium, International Finance, 2016.
31
See short description in Mackenzie, M., Ma , Financial Times, March 2013
32
For example, see Rachel, L., and Summers, L., On Falling Neutral Real Rates, Fiscal Policy, and the Risk of Secular Stagnation,
Brookings Papers on Economic Activity, 2019.
33
Caballero, R. J., Farhi, E. and Gourinchas, P-O., The Safe Assets Shortage Conundrum, Journal of Economic Perspectives, 2017.

Fiscal risks report 168

 Cost of public debt

Second, purchases of bonds by central banks have limited the quantity of safe assets
available to the non-bank private sector. Prior to the financial crisis, these purchases
largely consisted of reserve accumulation by emerging market central banks. But since
the financial crisis, quantitative easing by advanced economy central banks has
absorbed much of the new issuance of government bonds as of the final quarter of
2020, domestic central banks owned 26 per cent of general government debt in the
seven countries shown in Chart 4.6. Indeed, net of domestic central bank and foreign
official sector holdings, the supply of high-quality government bonds in private sector
hands has remained relatively constant as a share of global GDP, despite the large
increase in issuance by advanced economy governments. Absent these purchases by
the official sector, long-term interest rates would surely have been somewhat higher.

Chart 4.6: Holdings of selected high-quality government debt

70
Private sector holdings
60
Domestic central bank and foreign official sector holdings
50
Per cent of global GDP

40

30

20

10

0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
Note: General government debt for Australia, Canada, France, Germany, Japan, UK and USA
Source: IMF

4.35 As far as the outlook goes, this depends on not only future fiscal policies but also the
evolution of monetary policies. Fiscal policy across advanced economies has been
dramatically loosened to protect households and firms from the effects of the pandemic.
Governments in the US and elsewhere are contemplating further rounds of significant fiscal
stimulus to fuel the post-pandemic recovery. As output recovers and inflation pressures start
to build, central banks are likely to begin tightening monetary policy, including running
down some of their asset holdings, which all else equal would put upward pressure on
government bond yields. This could happen organically as bonds are not replaced when
they mature or through active sales programmes. But either way, the unwinding seems likely
to progress slowly.

169 Fiscal risks report

Cost of public debt

Box 4.3: The impact of the pandemic on real interest rates

Will the shock to the economy and public finances associated with the coronavirus pandemic
have any long-run impact on real interest rates? Jordà, Singh and Taylor (2020)a find that past
pandemics have indeed had long-lasting effects. Using Europeanb data stretching back to the
fourteenth century, they find that 20 years after a pandemic, the real interest rate was on
average around 1.5 percentage points lower (Chart B) and took around four decades to return
to pre-pandemic levels. However, the impact of pandemics on the real interest rate in the UK
was rather less (a decline of just 0.25 percentage points after 20 years).
The likely source of this effect is the fall in the labour force as a result of a higher number of
deaths, leaving a higher capital-labour ratio and a lower incentive to invest. That could be
augmented by increased savings by survivors.
This contrasts with the impact of wars, after which real interest rates have typically risen. That
most probably reflects the destruction of capital that typically occurs (especially in the wars of the
twentieth century) generating a post-war need to rebuild. In addition, governments have often
borrowed in order to fight wars, putting further upward pressure on interest rates.
Chart B: The impact of past pandemics and wars on interest rates
2.0
5.0

1.5
4.5

1.0
4.0

0.5
3.5

0.0
3.0
Per cent

-0.5
2.5

-1.0
2.0

-1.5
1.5
Wars
-2.0
1.0
Pandemics
-2.5
0.5
Shaded areas represent standard deviation bands
-3.0
0.0
0 5 10 15 20 25 30 35 40
Years since war or pandemic event ended
Source: Jordà, O., Singh, S. and Taylor, A. M., Longer-Run Economic Consequences of Pandemics, Covid Economics, 2020.

There are, however, reasons to think that the consequences of the coronavirus pandemic could
be different to previous pandemics. First, this pandemic has had only a limited effect on the
labour force, with total deaths being lower and more concentrated among the elderly. Second,
the large rise in borrowing accompanying this pandemic has been more like past wartime
episodes. That said, the response to the pandemic has largely filled the hole left by the
contraction in private sector spending, which is different from a war when fiscal expansions for
war spending and subsequent reconstruction place more pressure on available resources.c
a
Jordà, O., Singh, S. and Taylor, A. M., Longer-Run Economic Consequences of Pandemics, Covid Economics, April 2020.
b
Data for France, Germany, the Netherlands, Italy, Spain and the UK.
c
Hatzius, J., Daly, K., Struyven, D., Bhushan, S., and Milo, D., Inflation in the Aftermath of Wars and Pandemics, Goldman Sachs
Economics Research, March 2021.

Fiscal risks report 170

 Cost of public debt

Conclusions
4.36 The causes of the fall in global real interest rates have attracted much attention, but there is
no clear consensus in the literature about their relative importance. Chart 4.7 shows results
from several studies that have looked at the causes of falling global interest rates. Differing
time periods and definitions of the real interest rate mean that the size of the fall to be
explained varies across studies. But separate from that, it is clear that there is no consensus
on the relative importance of the different potential causes. Demography and falling
productivity figure most consistently in empirical studies. Since 2000, the shift in preferences
towards safer assets is likely to have been a factor pushing government bond yields below
returns on riskier assets, while purchases by central banks have helped to offset the upward
pressure coming from higher bond issuance. Other factors are likely to have played a part
too, though their precise contribution remains uncertain. This uncertainty is amplified by
longer-run studies such as that by Borio, Disyatat, Juselius and Rungcharoenkitkul who find
no robust relationship between real interest rates and any of the factors discussed above.34

Chart 4.7: Decomposition of the fall in global real interest rates

600
Demographics
Weaker trend growth/productivity
Preference for safe assets
400
Decline in relative price of capital goods
Inequality
Global savings glut
200
Other (including interactions and unexplained)
Higher government debt and other policies
Basis points

0
Total decline in real interest rate

-200

-400

-600

-800
Rachel & Smith Carvalho et al Lisack et al Rachel & Summers Eggertson et al Gagnon et al
(2015) (2016) (2017) (2019) (2019) (2021)
Note: The results of each paper are not directly comparable as not all cover the same set of factors. The exact time periods and
geographical areas covered by each paper also vary.
Source: Rachel, L., and Smith, T., Secular drivers of the global real interest rate, Bank of England Working Paper, 2015;
Carvalho, C., Ferrero, A., and Nechio, F., Demographics and real interest rates: Inspecting the mechanism, Federal Reserve Bank
of San Francisco Working Paper, 2016; Lisack, N., Sajedi, R., and Thwaites, G., Demographic trends and the real interest rate,
Bank of England Working Paper, 2017; Rachel, L., and Summers, L., On falling neutral real rates, fiscal policy, and the risk of
secular stagnation, Brookings Papers on Economic Activity, 2019; Eggertsson, G., Mehrotra, N., and Robbins, J., A model of
secular stagnation: Theory and quantitative evaluation, American Economic Journal: Macroeconomics, 2019; Gagnon, E.,
Johannsen, B., and Lopez-Salido, D., Understanding the new normal: The role of demographics, IMF Economic Review, 2021.

34
Borio, C., Disyatat, P., Juselius, M., and Rungcharoenkitkul, P., Why so low for so long? A long-term view of real interest rates, BIS
Working Paper 685, 2017.

171 Fiscal risks report

Cost of public debt

4.37 Uncertainty about the factors driving real interest rates in the past necessarily carries over
into the future. As already indicated, many of the potential drivers will continue to be in
place over the foreseeable future, though some may at least partly reverse. The factors
driving the balance between savings and investment (i.e. shifting the AA schedule in Figure
A of Box 4.2) represent mostly slow moving forces, such as demographics, which would
reverse and push real interest rates up only gradually. But there is perhaps scope for
sharper changes to occur due to changes in portfolio preferences (i.e. the PP schedule) or
monetary policy decisions.

4.38 Market expectations currently show a very gradual rise in real interest rates, with the level
remaining low historically, and this would allow the Government to continue to finance its
debt relatively cheaply. In our March 2021 EFO, we used market expectations for interest
rates on 5 February and, since then, the yield curve has risen as the economic outlook has
improved. Beyond 2025-26, the long-term economic determinants used on our fiscal
sustainability analysis assume that interest rates continue to rise to the point where they
exceed GDP growth rates by a small margin, taking the average gilt rate and Bank Rate to
steady-state levels of around 4 per cent.35 Given the uncertainty over the future path of real
interest rates, the remainder of the chapter explores how different scenarios for the evolution
of real interest rates would impact the UK public finances. We do this relative to a baseline
that is consistent with market expectations over the long term, since our long-term economic
determinants already assume that real rates revert to somewhat higher levels than is priced
into markets (and is in fact similar to our second scenario below).

The fiscal impact of higher global real interest rates

4.39 This section explores the fiscal implications of higher interest rates resulting from a rise in
global real interest rates (higher rates associated with higher inflation are discussed later).
Our focus is on the risk of higher, rather than lower, global interest rates, as this would be
more challenging for the Government to deal with. But it is worth noting that a further fall in
global real rates would present challenges of a different sort, particularly if it were
accompanied by disappointing productivity growth that weighed on the outlook for tax
receipts. In this world, the scope for monetary policy to support aggregate demand would
remain constrained by the proximity of the effective lower bound on Bank Rate, while the
efficacy of asset purchases in boosting demand is debatable when short- and long-term
rates are already so low.36 Fiscal policy would in this case need to be used even more
actively, placing more value on having sufficient fiscal space in reserve. Fortunately, the
continuation of low financing costs would also help preserve the fiscal space available.

The sensitivity of the public finances to interest rate changes

4.40 Before describing our scenarios, it is worth noting that the public finances are more sensitive
to a rise in interest rates than in the past as a result of the increase in the debt-to-GDP ratio
from the financial crisis and now the pandemic, which together have returned it to levels last

35
Most recently updated on 5 May 2021 on our website.
36
Vlieghe, G., , Bank of England, 23 April 2020.

Fiscal risks report 172

 Cost of public debt

seen in the early 1960s. In addition, by replacing longer-dated gilts with reserves that pay
Bank Rate, a by-product of quantitative easing has been to shorten the effective maturity of
consolidated public sector liabilities (i.e. consolidating all government debt and the Bank of
(APF)), so that the pass-through of changes in interest rates
happens faster (see Box 4.5, and also Box 4.1 in our March 2021 EFO).37

4.41 This greater sensitivity is illustrated in the left-hand panel of Chart 4.8. The total impact as a
share of GDP from a 1 percentage point rise in interest rates (that is the impact when all the
debt stock has moved to the higher rate) increases sharply during the financial crisis and
again in 2020 reflecting the increases in debt in those periods. The overall sensitivity in
2020 is three times that in 1998 (since the debt-to-GDP ratio has trebled from 33.9 to
100.2 per cent). The faster pace at which the increase passes through the debt stock is
illustrated by the proportion that responds in less than one year (proxied here by the stock of
Bank reserves used to buy gilts, Treasury bills, NS&I products and gilts with a residual
maturity of less than one year). The one-year impact was less than 0.1 per cent of GDP in
the decade up to 2008 but had risen nearly six-fold to over 0.5 per cent of GDP by the end
of 2020. More than half of this rise in the short-run sensitivity has come about as a by-
product of quantitative easing (since it has more than doubled the proportion of the much
higher debt that responds to interest rate changes within a year).

4.42 Commentators often employ the simple average (or mean) maturity of gilts as a summary
indicator of the speed of pass-through. This has risen from 10 to 15 years over this period
(the green line in the right-hand panel of Chart 4.8). But this offers a misleading picture of
the speed of pass-through of interest rate changes, and thus the immediacy of the fiscal
risks they pose, for three reasons:

First, it ignores other forms of debt issued by central government that have shorter
maturities, including Treasury bills and NS&I products. Taking these into account
shortens the mean maturity of this wider measure of debt to 13 years in 2020.

Second, one-third of the stock of gilts is now held in the APF,

financed by a corresponding issuance of central bank reserves that instead pay Bank
Rate which can change overnight (as explained in Box 4.1 of our March 2021 EFO).
This further reduces the mean maturity of the consolidated liabilities of the public
sector in 2020 to nine years.

Third, the mean maturity is itself a misleading guide to the speed of pass-through of
interest rate changes into the public finances. That is because the mean can be skewed
by the presence of a relatively small volume of very long maturity bonds. This is the
case for the UK, where 27 per cent of outstanding gilts held by the private sector have
a maturity of over 15 years. But the median maturity of the consolidated liabilities of
the public sector is in fact only about two years at the end of 2020, while 45 per cent
of the liabilities have an effective maturity of less than one year. As a result, much of

37
It is also worth noting that because reserves do not have to be rolled over they are effectively floating rate perpetuities this
shortening of effective maturity in terms of the speed of interest rate pass-through is accompanied by reduced, rather than increased,
refinancing risk.

173 Fiscal risks report

Cost of public debt

the impact of higher interest rates on the public finances now actually comes through
quite rapidly.

4.43 The median maturity of these consolidated public sector liabilities is therefore a more
suitable summary measure when considering the short-term fiscal risks posed by interest
rate changes (the red line in Chart 4.8), as it represents a direct measure of the time it takes
for half of the full effect of a rise in rates to be feed through to interest payments.

Chart 4.8: Sensitivity of interest payments to a rise in interest rates

1.2 26
Rise in interest payments from a one Maturities
24
percentage point rise in interest rates Mean of gilts
1.0 22
Total impact Mean of gilts, Treasury bills and NS&I
20
of which within one year Mean of public sector liabilities (net of APF)
0.8 18
Median of public sector liabilities (net of APF)
Per cent of GDP

16
14
Years
0.6
12
10
0.4
8
6
0.2 4
2
0.0 0
1998 2001 2004 2007 2010 2013 2016 2019 1998 2001 2004 2007 2010 2013 2016 2019
Note: Consolidated public sector liabilities are proxied here by the stock of Bank reserves, Treasury bills, NS&I products and gilts
net of those held in the APF. The total impact of a one percentage point rise in interest rates is on consolidated public sector
liabilities. The impact within one year is on liabilities with a maturity of less than 1 year (gilts net of APF holdings with a remaining
maturity of under one year, Bank reserves, Treasury bills and NS&I). The median shows the year in which half of the outstandi ng
public sector liabilities would be impacted by a change in interest rates.
Source: Bank of England, Heriot-Watt/Faculty and Institute of Actuaries Gilt Database, ONS, OBR

Baseline assumptions
4.44 Turning to how this affects our scenarios for real interest rates, our baseline assumes a
profile for interest rates similar to that anticipated by market participants at the time of our
March 2021 EFO. Both the baseline and the scenarios extend to 2050-51 in order to allow
us to evaluate the medium- and long-term implications of higher interest rates. The key
assumptions in the baseline are as follows:38

Nominal GDP follows our medium-term forecast from the March 2021 EFO and is
constant at 3.9 per cent thereafter (in line with the average rate in steady state in our
long-term economic assumptions).

CPI inflation follows our March forecast up to 2025-26 and is constant at the 2 per
cent target thereafter.39

38
An additional assumption relates to stock-flow adjustments (changes in debt not accounted for by the primary balance or debt interest),
which follow our March 2021 EFO forecast to 2025-26. Part of this adjustment comes from
Funding Scheme. After 2025-26, we assume the scheme is run down over five years. We assume other stock-flow adjustments are 0.7 per
cent of GDP in each year after 2025-26, in line with the later years of our March forecast.
39
For simplicity, we assume that the GDP deflator, RPI, CPI all move together in the scenarios.

Fiscal risks report 174

 Cost of public debt

The primary balance follows our medium-term forecast from the March 2021 EFO to
2025-26. After this, spending and receipts remain a constant share of GDP, which
means the primary deficit remains at 1.8 per cent of GDP. (This abstracts from
pressures due to ageing and other factors that are covered in the FSRs).

Bank Rate rises in line with our March forecast, based on market expectations on 5
February. Beyond 2029-30, Bank Rate remains constant at 1.1 per cent, which is
assumed to be consistent with the underlying equilibrium real interest rate in the
baseline and with meeting the inflation target.

The average gilt rate40 increases in line with our March forecast and remains at 1.3
per cent from 2029-30 onwards. We assume that 7 per cent of gilts are refinanced
each year (in line with the average between 2020-21 and 2025-26), which means any
changes in average gilt rates feed through gradually to the G ctive
interest rate. For simplicity, we assume that all new government debt issued is in the
form of gilts (88 per cent in conventional gilts and 12 per cent in index-linked gilts in
line with the financing assumption in our March 2021 EFO)41.

The APF follows our March forecast and thereafter the stock of reserves is kept constant
in nominal terms, paying the prevailing Bank Rate in interest costs.

For other interest payments and receipts, such as Treasury bills and NS&I products,
interest rates are assumed to move in line with either Bank Rate or the average gilt
rate.

Alternative global real interest rate scenarios

4.45 We consider two alternative scenarios in which interest rates rise as a result of higher global
real interest rates. Reflecting our earlier discussion of the outlook for global real rates, this
rise is assumed to happen relatively slowly:

In our , real interest rates and real GDP growth rise in

tandem, for instance reflecting a recovery in productivity growth.

In our , the interest rate rises without a corresponding rise

in growth, for instance reflecting a shift in portfolio preferences away from bonds. This
-term economic
.

4.46 In both cases, the Bank of England is assumed to correctly diagnose what is happening and
so raises Bank Rate in line, leaving inflation totally unaffected. Both scenarios also assume
that there is no change in the size of the APF (Box 4.5 discusses the different pressures that
could arise if the Bank opted to run it down). For the first two years of each scenario, we

40
This is the marginal cost of new gilt issuance. It is the weighted average of yields across all maturities.
41
We assume that new gilts are issued with coupons such that they are sold at par.

175 Fiscal risks report

Cost of public debt

also assume that non-welfare spending is fixed in cash terms and receipts move by more
than any change in nominal GDP (due to fiscal drag), but beyond that public spending and
receipts move one-for-one with nominal GDP. This allows us to highlight the mechanics
through which higher interest rates affect the public finances. The specific assumptions for
the path of R and G in the two scenarios are shown in Chart 4.9 and summarised below:

In the higher R and G scenario, average gilt rates and Bank Rate gradually rise above
the baseline over the next decade to settle 2.5 percentage points higher. This reverses
approximately half of the fall in long-term bond yields over the past 20 years. But in
historical terms, gilt rates and Bank Rate still finish at relatively low levels, at 3.8 and
3.6 per cent respectively. Real GDP growth rises in line with real interest rates,
reaching 2.5 percentage points above our baseline by 2032-33. This would represent
a large pick-up in real GDP growth to a rate last seen in the late 1980s. Inflation is
unchanged, so nominal GDP growth rises in line with real GDP growth.

In the higher R scenario, we assume the same increase in interest rates as in the first
scenario but leave real and nominal GDP growth unchanged from the baseline.

Fiscal risks report 176

 Cost of public debt

Chart 4.9: Higher global real interest rate scenarios: key assumptions
7
Average gilt rate Baseline
6
Higher R and G
5
Higher R
4
Per cent

0
2000-01 2005-06 2010-11 2015-16 2020-21 2025-26 2030-31 2035-36 2040-41 2045-46 2050-51
7
Bank Rate Baseline
6
Higher R and G
5
Higher R
Per cent

0
2000-01 2005-06 2010-11 2015-16 2020-21 2025-26 2030-31 2035-36 2040-41 2045-46 2050-51
10
Nominal GDP growth
8

4
Per cent

0
Baseline
-2
Higher R and G
-4
Higher R
-6
2000-01 2005-06 2010-11 2015-16 2020-21 2025-26 2030-31 2035-36 2040-41 2045-46 2050-51

Note: Average gilt rate is the average yield of gilt issuance before 2012-13
Source: Bank of England, DMO, ONS, OBR

4.47 Chart 4.10 shows the fiscal results from these two scenarios. In interpreting these it is
important to note that in the higher R and G scenario, the growth-corrected interest rate
falls somewhat rather than remaining unchanged. This is because, while the growth rate
increases immediately, it takes time for higher market interest rates to feed through to the

177 Fiscal risks report

Cost of public debt

average interest rate paid on all the G . Over time the growth-corrected
interest rate then converges back to that in the baseline.

4.48 Borrowing rises compared to the baseline, reaching 5.1 per cent of GDP in 2050-51
compared to 2.9 per cent in the baseline. There is a small initial benefit to primary
borrowing from the impact of fiscal drag on tax revenues and our assumption that non-
welfare spending is fixed in cash terms for two years. But thereafter we assume the
Government spends the proceeds of stronger nominal GDP growth so primary borrowing
returns to the same level as in the baseline. Borrowing instead increases on the back of
higher net interest payments, which by 2050-51 are more than three times the 1.0 per cent
of GDP in the baseline, reaching 3.3 per cent of GDP a level last seen in 1985-86.
Interest costs as a proportion of revenue are more than three times higher, reaching 8.6 per
cent in 2050-51 compared to 2.7 per cent in the baseline.

4.49 Throughout the scenario, and despite higher borrowing, the debt-to-GDP ratio is lower than
in the baseline, primarily due to the more favourable growth-corrected interest rate. By
2050-51, PSND is 6.4 per cent of GDP below baseline. However, the debt-to-GDP ratio
does not return to its pre-pandemic level by the end of the scenario.

4.50 In the higher R scenario, nominal GDP growth is unchanged but the higher Bank Rate and
average gilt rate rapidly feed through to the effective interest rate the Government pays on
its debt. The growth-corrected interest rate therefore rises compared to the baseline,
although not quite enough to push it into positive territory (Chart 4.10).

4.51 Higher interest rates mean that net interest payments are five times higher than the baseline
at 5.0 per cent of GDP in 2050-51. Net interest payments as a proportion of revenue rise
from 2.7 in the baseline to 13.2 per cent by the end of the scenario, their highest since
1946-47. Higher spending on interest payments pushes borrowing above the baseline
throughout the scenario, reaching 6.8 per cent of GDP in 2050-51 compared to 2.9 per
cent in the baseline.

4.52 Debt is also significantly above the baseline due to the less favourable growth-corrected
interest rate. Debt rises to 139 per cent of GDP by 2050-51 almost 43 percentage points
above the baseline. The higher growth-corrected interest rate in this scenario means that to
stabilise debt, the Government would need to reduce the primary deficit in 2029-30,
lowering it by 0.9 per cent of GDP to 0.9 per cent would be sufficient; by 2050-51, the
adjustment would need to be 2.0 per cent of GDP approximately equivalent to the size of
the defence budget in 2020-21.

Fiscal risks report 178

 Cost of public debt

Chart 4.10: Higher global real interest rate scenarios: key outputs
10 6
Effective interest rate less nominal Net interest payments
8
GDP growth (R-G) 5
6

4 4

Per cent of GDP

2
Per cent

3
0

-2 2

-4
1
-6

-8 0
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
20 160
PSNB PSND
140
15
120
Per cent of GDP
Per cent of GDP

10 100

80
5

60
0 Higher R and G
40
Higher R
Baseline
-5 20
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
Source: ONS, OBR

Chart 4.11: Higher global real interest rate scenarios: contributions to differences in
debt-to-GDP ratios from the baseline
45 45
Higher R and G Higher R
40 40
Per cent of GDP difference from baseline

Per cent of GDP difference from baseline

Growth-corrected interest rate (R-G)

35 35
Stock-flow adjustments
30 30
Primary deficit
25 25
PSND
20 20
15 15
10 10
5 5
0 0
-5 -5
-10 -10
2022-23 2030-31 2038-39 2046-47 2022-23 2030-31 2038-39 2046-47
Source: OBR

179 Fiscal risks report

Cost of public debt

4.53 These scenarios envisage a gradual rise in interest rates relative to a baseline based on
market expectations at the time we finalised our March 2021 EFO. Part of this rise has
already crystallised because the yield curve has risen since our March forecast. The average
gilt rate has risen by almost 50 basis points and market expectations for Bank Rate over the
next five years have risen by 30 basis points. These changes would raise debt interest costs
by around £7 billion in 2025-26, almost half of the increase in our higher R scenario.

Conclusions
4.54 The relatively benign scenario of a gradual increase in growth alongside interest rates gives
a modest reduction in debt, although it does not return to its pre-pandemic level relative to
GDP by the end of our scenario period in 2050-51. But even in this scenario, net interest
payments reach more than three times the level in the baseline. Higher interest rates in the
absence of higher GDP growth deliver a worse outcome for the public finances, with debt
and borrowing climbing throughout the scenario. By 2050-51, the debt-to-GDP ratio
reaches its highest level since 1954-55.

4.55 There are a several caveats to these scenarios worth mentioning. First, our simulations do
not attempt to capture the full range of economic and fiscal effects from the changes in
asset prices as bond yields rise. This could affect financial stability, for example, if they
happened abruptly. Second, forward-looking financial markets could mean the fiscal
benefits in the higher R and G scenario are more limited than shown above. We assume
interest rates and growth gradually rise together, but investors could demand higher interest
rates in anticipation of the pick-up in growth, thereby reducing the initial fiscal benefits.

Fiscal risks report 180

 Cost of public debt

The fiscal impact of higher inflation

4.56 Although interest rates could rise because of a partial reversal of the factors that have driven
them down over the past three decades, it is also possible that real rates could remain
unchanged but nominal rates rise as a consequence of higher inflation or inflation
expectations. Prompted in part by continued expansionary US fiscal policies,42 but also
reflecting the continued accommodative monetary policies in many jurisdictions, there has
been growing debate as to whether a resurgence of inflation is in prospect. There have also
been suggestions that governments might welcome a period of higher inflation as a means
of reducing the real value of their outstanding debt.43 In this section, we therefore consider
the impact on the public finances of a rise in yields that is associated with higher inflation.
As inflation is ultimately a domestic phenomenon for a country with its own currency and a
floating exchange rate, this should be thought of as reflecting UK policy choices or other
UK-specific factors. But it is possible, of course, that there may a general tendency to higher
inflation across multiple jurisdictions, reflecting the operation of the same factors in other
countries.

4.57 Again, we consider two scenarios. In the first there is a burst of domestically-generated
inflation that we assume requires a temporary rise in Bank Rate to bring inflation back to
target. In the second there is a more persistent rise in inflation, which could be associated
either with sustained failure to meet the inflation target or the adoption of a higher one. In
both, we continue to assume that non-welfare spending is fixed in cash terms for just the
first two years and beyond that rises in line with inflation, reflecting pressure for higher pay
and to maintain the supply of government services. We also retain the assumption that
receipts move with nominal GDP initially more than one-for-one due to fiscal drag, and
subsequently one-for-one in line with historical evidence on longer-term tax buoyancy.44

Temporary rise in inflation

4.58 A temporary rise in inflation could result from any number of shocks. In the current context it
is perhaps easiest to think of it as resulting from an overestimation of the margin of spare
capacity during the recovery from the pandemic that results in both higher inflation and
higher inflation expectations. This necessitates a tightening of monetary policy to bring
inflation durably back to the target, as this would not be the sort of inflation shock that the
Bank could simply look through. Tighter monetary policy, in turn, acts as a temporary drag
on GDP growth, bringing aggregate demand back into line with supply. In this scenario:

Inflation rises sharply in 2022-23, hitting 5 per cent (3 percentage points above both
our baseline and the target) the following year.

42
For example, see Summers, L., Comments to Federal Reserve Bank of Atlanta conference, May 2021.
43
For example, see discussion in Bank for International Settlements, Annual Economic Report, June 2020.
44
See Table 1 in Belinga, V., Benedek, D., de Mooij, R. and Norregaard, J., Tax buoyancy in OECD countries, IMF Working Papers No
14/110, International Monetary Fund, June 2014.

181 Fiscal risks report

Cost of public debt

The Bank of England reacts by raising Bank Rate to 4 per cent in 2022-23 (3.9
percentage points above baseline).45 Average gilt rates also rise to reflect the higher
path for Bank Rate.

Higher interest rates lead to weaker GDP growth over 2022-23 and 2023-24 (0.5
percentage points below the baseline in both years). There is no impact on potential
output, so the output gap widens over these two years.

Beyond the near term, inflation subsequently falls back, returning to target after four
years, while GDP growth, Bank Rate, and average gilt rates all also return to their
baseline paths over a similar timeframe.

4.59 The burst of inflation initially improves the primary balance due to fiscal drag lifting receipts
and departmental expenditure falling as a share of GDP due to being fixed in cash terms
for two years. The primary balance subsequently returns to baseline as the Government
increases cash spending and adjusts tax thresholds to account for the impact of inflation.

4.60 Debt interest payments rise immediately, underscoring the growing sensitivity of the debt
stock to changes in both inflation and interest rates. Higher inflation has a direct impact on
the interest payments on the stock of index-linked gilts, pushing interest costs up by £9
billion (0.4 per cent of GDP) in 2022-23. Similarly, the hike in Bank Rate leads to an
immediate increase in the interest paid on reserves of £34 billion (1.4 per cent of GDP),
reducing remittances from the APF to the Treasury by a corresponding amount (Box 4.5 has
a fuller explanation). Finally, higher average gilt rates raise interest costs more gradually.
This slower pace is because the Government only pays the interest rate prevailing in the
market on new gilts, issued either to finance the deficit or refinance the 7 per cent of gilts
assumed to mature each year (Chart 4.13 shows the breakdown).46 In the long run, there is
no change in annual net interest costs because this is a transitory shock and interest rates
return to their baseline levels.

4.61 The effect of a temporary burst of inflation on the debt stock is quite modest in both the
short and the long run. The debt-to-GDP ratio initially falls more quickly than in the
baseline, due to a lower primary deficit and more favourable growth-corrected interest rate
(Chart 4.13). But the fiscal benefit is quite small as interest costs rise quickly (particularly on
index-linked gilts and on reserves as Bank Rate is increased) and, by assumption, starting in
the third year of the scenario the Government increases cash spending to account for the
impact of higher inflation. By 2050-51, debt is 2 per cent of GDP below the baseline at 95
per cent of GDP, but is still over 10 per cent of GDP above the pre-pandemic level.

45
Estimated using the model described in Working paper No.4: A small model of the UK economy, OBR, July 2012.
46
We assume that gilts are issued at par, in line with general DMO practice.

Fiscal risks report 182

 Cost of public debt

Chart 4.12: Higher inflation scenarios: key outputs

10 7
Effective interest rate less Inflation (GDP deflator)
8 nominal GDP growth (R-G) 6

6 5

4 4

Per cent of GDP

2 3
Per cent

0 2

-2 1

-4 0

-6 -1

-8 -2
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51
20 120
PSNB PSND
110

15 100

90
Per cent of GDP
Per cent of GDP

10 80

70
5 60

50
Temporary inflation shock
0 40
Persistent rise in inflation
30
Baseline
-5 20
2000-01 2010-11 2020-21 2030-31 2040-41 2050-51 2000-01 2010-11 2020-21 2030-31 2040-41 2050-51

Source: ONS, OBR

Chart 4.13: Temporary inflation shock impact on net interest payments and PSND
2.5 2
Net interest payments PSND
Per cent of GDP difference from baseline

1
Per cent of GDP difference from baseline

2.0
Inflation
Bank Rate 0
Average gilt rate
1.5 -1
Real GDP
Net interest payments
-2
1.0
-3
0.5 -4

-5 Stock-flow adjustments
0.0
Growth-corrected interest rate (R-G)
-6
Primary deficit
-0.5 PSND
-7
2022-23 2030-31 2038-39 2046-47 2022-23 2030-31 2038-39 2046-47
Source: OBR

183 Fiscal risks report

Cost of public debt

Persistent rise in inflation

4.62 Our second inflation scenario examines the fiscal impact of a persistent increase in inflation
and inflation expectations to 2 percentage points above the baseline. Within the confines of
the existing monetary policy framework, that might result from a temporary inflation
overshoot that became embedded in actual and expected inflation and was expressly
accommodated through the open letter process.47 But it would also be consistent with the
suggestions from some quarters that the adoption of a higher inflation target would give
more room for central banks to lower their policy rates before hitting their effective lower
bound.48

4.63 In this scenario:

CPI inflation rises to 4 per cent over three years and remains at that rate.49 The process
takes several years because of frictions in adjusting prices and wages.50 Bank Rate
rises in step with inflation, leaving the real short-term interest rate unchanged.

Gilt rates react immediately to higher expected future paths of inflation and Bank Rate.
But we assume that gilt rates rise by 3 percentage points rather than the 2 percentage
point increase in the inflation expectations,51 because investors fear there might be
greater willingness to tolerate even higher rates of inflation in the future.

4.64 We assume that the economy adjusts smoothly to the persistently higher path for inflation so
there is no impact on real GDP or the output gap. If that were not so, then there would be
secondary impacts on borrowing and the debt-to-GDP ratio. A persistent inflation shock
raises overall borrowing in every year of the forecast. The primary deficit initially falls as
inflation rises, mostly due to nominal government spending being fixed for the first two
years. However, overall PSNB still rises because higher net interest costs outweigh the
impact of a lower primary deficit. The higher net interest costs come from three sources: the
direct impact of higher inflation on the cost of index-linked gilts; the impact of higher Bank
Rate on interest paid by the APF; and the impact of a higher average gilt rate on
conventional gilts (Chart 4.14). The first two feed through immediately, but the third feeds
through more slowly as existing gilts mature and new gilts are issued. Net interest costs rise
to 4.4 per cent of GDP by 2050-51 compared to 1.0 in the baseline (this takes them from
2.7 per cent of revenue in the baseline to 11.6 per cent in the scenario in 2050-51). This
raises borrowing to 6.2 per cent of GDP in 2050-51 compared to 2.9 per cent in the
baseline.

47
If CPI inflation deviates from the 2 per cent target by more than 1 percentage point, the Governor of the Bank of England is required to
write to the Chancellor explaining why and what will be done about it. The Chancellor is required to respond and could in theory set out
ces for how quickly the Bank should aim to address the deviation given the trade-offs involved. See, for
example, the discussion in Carney, M., Lambda, 16 January 2017.
48
Rethinking macroeconomic policy, Journal of Money, Credit and
Banking, 2010.
49
This is consistent with the proposal in , G., and Mauro, P., (op. cit.).
50
In the charts in this chapter, we show the GDP deflator rather than CPI (which the Bank of England targets). The GDP deflator settles at
slightly higher than 4 per cent but still 2 percentage points higher than the baseline, which is what is important for the scenario
calculations.
51

Fiscal risks report 184

 Cost of public debt

4.65 In this scenario, persistently higher inflation does nothing to reduce the debt-to-GDP ratio
over the long run. Debt initially falls marginally below the baseline due to the initial impact
of lower primary borrowing, but it ends up above the baseline by the end of the scenario as
higher interest rates work their way through the debt stock. In this scenario, the growth-
corrected interest rate is slightly less favourable than in the baseline. Although nominal
growth is lifted by higher inflation, the effective interest rate on government debt rises by
more due to the assumed inflation risk premium on average gilt rates. This means debt rises
back towards our baseline before exceeding it in the later years of the scenario. It reaches
107 per cent of GDP in 2050-51 (10 per cent of GDP above the baseline). This is more
than explained by the assumed 1 percentage point inflation risk premium on gilt rates,
which adds 12 per cent of GDP to debt in 2050-51.

Chart 4.14: Persistent rise in inflation impact on net interest payments and PSND
4.0 14
Net interest payments PSND
12
Per cent of GDP difference from baseline
Per cent of GDP difference from baseline

3.0 Stock-flow adjustments

10
Growth-corrected interest rate (R-G)
8
2.0 Primary deficit
6
PSND
1.0 4

2
0.0
Inflation 0
Bank Rate -2
-1.0
Average gilt rate
-4
Net interest payments
-2.0 -6
2022-23 2030-31 2038-39 2046-47 2022-23 2030-31 2038-39 2046-47
Source: OBR

4.66 The impact of rising yields on the market value of gilts is an additional factor not considered
in this scenario. For conventional gilts currently in issue, a 100 basis point rise in yields
would lower the average market value by 12 per cent, so the scenario would be consistent
with them falling by around a third. This could adversely affect financial stability.52

Conclusions
4.67 Both scenarios suggest that inflation is not a very effective way to reduce the debt-to-GDP
ratio in the current circumstances. A temporary burst of inflation has only a modest impact
on the debt-to-GDP ratio, which is mostly achieved through a temporary squeeze on real
spending. A persistent increase in inflation leads to a medium-term improvement in the debt
position as inflation erodes the real value of the nominal debt in issue. But the impact is
muted by the share of index-linked debt (23 per cent of gilts in 2020-21, up from 14 per
cent in 1989-90) and the shortening of the effective maturity of debt due in part to
quantitative easing. In the long run, there is actually a rise in the debt-to-GDP ratio due to
the assumption of a higher inflation risk premium on gilts, which pushes interest payments
from 1 to over 4 per cent of GDP, up to a level not seen since 1947-48.

52
See discussion in Bank for International Settlements, Financial Stability implications of a prolonged period of low interest rates, July
2018.

185 Fiscal risks report

Cost of public debt

Loss of investor confidence

4.68 Our scenarios so far have explored the fiscal implications of relatively modest, and often
gradual, rises in the cost of government borrowing. However, history is replete with
instances where governments experienced sudden and steep rises in the rates they had to
pay to borrow, sometimes even ending in complete loss of investor confidence and market
access. In such circumstances, access to an emergency loan from the International Monetary
Fund (or similar official lender) may buy time, but a significant fiscal consolidation will
usually be necessary to return the public finances onto a sustainable path and restore
investor confidence and market access. The remainder of this chapter reviews some of the
evidence regarding such government debt crises before exploring an extreme tail-risk
scenario in which the UK is subjected to one.

Government borrowing costs in a debt crisis

4.69 Governments are likely to find it more expensive to borrow
when investors lose confidence in their ability, or willingness, to honour their debt
commitments. Factors that erode investor confidence include not only fears of outright
default but also concerns that the government may take actions that have a similar effect,
such as reducing the real value of those debt obligations through higher inflation.
Governments rarely choose freely to repudiate their debt obligations, as that could result in
them finding it much more expensive to borrow in the future at least for a period.53
Instead, they are usually forced into default or to resort to the IMF or other multilateral or
bilateral loans by a combination of rising borrowing costs and difficulties in delivering the
rapid adjustments to the primary balance needed to offset those rising interest payments. As
a result, debt crises often feature adverse feedback loops, with higher borrowing costs
worsening the fiscal position and the worsening fiscal position leading to higher borrowing
costs.

4.70 Chart 4.15 shows the behaviour of UK gilt yields around the time of its 1976 crisis, the last
time the UK had to seek external assistance from the IMF. The chart also shows the yields on
government bonds during some more recent government debt crises in other advanced
economies. In each case, yields rose substantially in just a matter of months, illustrating how
rapidly financing conditions can deteriorate.

4.71 The UK crisis in 1976 was on the surface a balance of payments crisis, though associated
fundamentally with unsustainable fiscal and monetary policies illustrating the range of
factors that can combine to create a debt crisis. This combination of factors is reflected in
the flatter profile of interest rate rises in Chart 4.15 longer-term issues of high inflation,
and the after-effects of a recession and the oil crisis left the deficit high, and gilt rates in
1975 already stood 7 percentage points higher than their 1960s average. Against this
backdrop, from early 1976 market participants believed that sterling devaluation was

53
See the discussion of the impact of a history of default on borrowing costs in Cruces, J., and Trebesch, C., Sovereign Defaults: The Price
of Haircuts, American Economic Journal: Macroeconomics, 2013, and earlier work by Borensztein, E., and Panizza, U., The Costs of
Sovereign Default, IMF Working Paper, October 2008.

Fiscal risks report 186

 Cost of public debt

inevitable in view of -term rates rose by over 3

percentage points within seven months but, more damagingly, investors came to regard

Eventually, an IMF loan of $3.9 billion (1.2 per cent of UK GDP in 1976) was necessary, the
price of which was the implementation of politically unpalatable cuts in public spending.54

Chart 4.15: Long-term bond rates in selected government debt crises

35

One year before peak

30
Long-term interest rates on government debt

Six months before peak

25
Peak during crisis
20

15

10

0
UK - 1976 crisis Ireland - 2010 crisis Iceland - 2008 crisis Greece - 2009 crisis
Source: OECD, OBR

4.72 The other three episodes in Chart 4.15 all date from the financial crisis, when the cost of
stabilising banking sectors against a backdrop of severe recessions sharply worsened fiscal
positions in affected countries. Greece is perhaps the most notable, with large revisions to
the pre-crisis fiscal accounts precipitating a crisis that led to the most significant bail-out in
Europe. Revelations over the course of 2009 that the budget deficit was far higher than
realised, and further deterioration in the fiscal position as a result of the financial crisis, led
investors to think that both default and/or exit from the euro might be necessary. This
caused interest rates to rise dramatically eventually peaking at 29 per cent in February
2012. This was ultimately only resolved by a mixture of external support from the EU and
IMF, and a commitment from the European Central Bank to maintain the integrity of the
euro by purchasing Greek and other eurozone sovereign debt through its Outright Monetary
Transactions programme. This was followed by a restructuring of private sector holdings of
Greek debt in 2012, which reduced the face value of these private holdings 100
billion.55

4.73 The cases of Iceland and Ireland are notably different, insofar as neither entered their crises
with weak fiscal positions. But in both, the banking system was so large relative to the
economy that the costs of rescuing it implied a sudden and very large transfer of liabilities

54
The National Archives, Sterling devalued and the IMF loan, accessed May 2021.
55
For more discussion see: Zettelmeyer, J., Trebesch, C. and Gulati, M., The Greek Debt Restructuring: An Autopsy, August 2013.

187 Fiscal risks report

Cost of public debt

from private to public sector increasing gross debt levels by 24.5 per cent of 2010 GDP in
Ireland between 2007 and 2010, and by 42.8 per cent of GDP in Iceland over the same
period.56 In Ireland, the announcement of the
of was followed by a doubling in interest
rates over the next ten months. In Iceland, the announcement of the nationalisation of the

costs, with interest rates rising nearly 6 percentage points from September to October.

Box 4.4: Long run drivers of UK government debt

The history of UK government debt can be
which long periods where the debt-to-GDP ratio is broadly stable or gradually falling are
interspersed with occasional large increases in response to major shocks (Chart 4.1). Since 1900
(and including our March 2021 forecast), there have been 21 years in which debt rose by more
than 6 per cent of GDP (Chart C); of these, 12 were associated with the world wars and the
remaining nine came during the contractionary policy in the 1920s (two), the Great Depression
(two), the financial crisis (three), and the coronavirus pandemic (two).
The two world wars were by far the most significant events. The first saw five years of debt-to-
GDP increases averaging 23 percentage points a year, while the second saw seven consecutive
years of rising debt at an average of 15 percentage points a year. While the total increase in
debt during the world wars was far greater than during the pandemic, they are similar in being
exogenous shocks that led to large, though temporary, increases in primary spending, partially
offset by falling nominal interest rates.
Chart C: Distribution of changes in the UK debt-to-GDP ratio
30

25

20
Number of years

15

10

0
Less than -10 to -8 -8 to -6 -6 to -4 -4 to -2 -2 to 0 0 to 2 2 to 4 4 to 6 6 to 8 8 to 10 More
-10 than 10
Change in debt-to-GDP ratio
Source: Bank of England, ONS, OBR

56
IMF, Fiscal Transparency, Accountability, and Risk, August 2012.

Fiscal risks report 188

 Cost of public debt

After a crisis, governments usually seek to rebuild fiscal space in order to be able to respond to
the next crisis. For crises driven by temporarily higher spending such as wars and the pandemic
a rapid improvement in the primary balance should be possible simply by returning
expenditure closer to pre-crisis levels once the need for the temporary rise has passed. But
returning debt to pre-crisis levels can be the work of many decades.
The UK successfully brought debt down following the second world war. After 1946-47 the debt-
to-GDP ratio fell for 27 consecutive years and by 206 per cent of GDP. Of this, 127 percentage
points was achieved in the first decade. More than half the fall was achieved by a persistently
negative growth-corrected interest rate. In part this was the result of interest rates being held
,a but the Government
also ran relatively large primary surpluses, particularly in the early post-war period when the
civilian workforce (and therefore tax revenues) was expanding rapidly. Later in the period,
particularly from the late 1960s to the 1980s, persistently high (and sometimes unanticipated)
inflation also helped to erode the real value of the Government debt stock at a time when
nominal interest rates were still subject to administrative control.
Of the various strategies that contributed to the post-war debt reduction, the most desirable for
society and bondholders alike would clearly be sustained higher real GDP growth (consistent
with our ) but this has proved extremely difficult to achieve in the post-
financial crisis period. Financial repression would be more difficult to achieve in an era of open
capital markets and independent central banks and financial regulators. And our ersistently
higher inflation scenario suggests that a period of higher inflation may no longer offer an
effective way of reducing the debt-GDP ratio, especially if it results in a higher inflation risk
premium.
a
See for example Reinhart and Sbrancia, The liquidation of government debt, 2011.

What affects the risk premium on government debt during a crisis?

4.74 Of course, no two crises are alike and their fiscal consequences depend in part on how
governments respond. So the historical and international experiences discussed above may
not capture the sorts of risks the UK might face were it subject to one. However, empirical
studies of interest rate differentials on government debt and sovereign debt crises point to
several factors that drive the risk premium on government bonds.57 When investors have full
confidence in the creditworthiness of a government, the quantitative importance of these
factors may not be that great. Indeed, such bonds may well safe
haven (see the discussion of fiscal space in Box 1.1). But
these factors are likely to come to the fore when the creditworthiness of the Government is
in doubt.

4.75 There are several potential drivers of risk premia on government bonds at such times. We
begin with the outlook for government deficits and debt. Other things equal, higher debt
paths could be expected to put upward pressure on yields because they increase the risk of

57
For a fuller discussion of risk premia, see Gürkaynak, R. and Wright, T., Macroeconomics and the Term Structure, Journal of Economic
Literature, June 2012.

189 Fiscal risks report

Cost of public debt

future capital losses, either because yields continue to rise in the future or because of the
heightened risk of some sort of default. While UK public borrowing has increased
dramatically as a result of the pandemic and public debt has risen sharply as a result, the
former should fall sharply as the pandemic recedes and the UK is in the middle of the pack
of advanced economies as far as its debt-GDP ratio is concerned (see Chart 4.16).

4.76 Second, investor perceptions on the risk of default are also related to the profile of the
G financing needs, which depend not only on the budget deficit but also on the
quantity of maturing debt that needs to be rolled over. A high volume of short-term issuance
makes a government more vulnerable to funding problems and shortens the time available
to get the public finances in order. So, while funding at short maturities is typically cheaper,
it also leaves the Government more at risk. Earlier in this chapter, we explored the effect of
quantitative easing on interest rate sensitivities via the effective shortening of the maturity of
the public debt. However, from a funding perspective, central bank reserves do not have to
be refinanced; they are, in effect, a floating rate perpetuity. What matters instead is the total
new debt that the Government needs to place with private buyers. Chart 4.16 shows
projected financing requirements and debt burdens for advanced economies in 2021.
Despite a high stock of debt, the UK has a lower financing requirement compared to other
advanced economies with similar debt burdens. This reflects the relatively long average
maturity of UK gilts a factor that reduces the G financing risk.58

Chart 4.16: General government net debt and gross financing needs, 2021
75
Gross financing need, as a proportion of GDP

Japan
60

United States
45
G7 average
Adv. Economy average
30
Italy
Canada UK
France
Germany
15

0
0 30 60 90 120 150 180
General government net debt, as a proportion of GDP
Note: Data shown for advanced economies, excluding Greece, Cyprus, Estonia, Hong Kong, Israel, Latvia, Luxembourg, Malta,
Norway, Singapore and the Slovak Republic due to data availability.
Source: IMF

58
Note that Bank reserves never mature and so do not need refinancing. For this reason, a maturity measure only including gilts is
perhaps rather better as an indicator of financing risk.

Fiscal risks report 190

 Cost of public debt

4.77 Third, perceptions of government debt risk also depend on wider pressures on the public
sector balance sheet, including the stock of debt-like obligations, such as public sector
pensions, as well as the availability of assets that could be liquidated if required to meet
government financing needs. IMF research suggests that an improvement in an advanced
10 per cent of GDP on average lowers its bond yields
by just under 10 basis points.59 Chart 4.17 shows the net worth position of selected
advanced economies. Among these countries, the UK
public sector pension liabilities, and paucity of financial and non-financial assets, place it at
the bottom of the league table.

Chart 4.17: General government net worth for selected advanced economies
700
Total assets
600
Total liabilities
500
Net worth
400

300
Per cent of GDP

200

100

-100

-200

-300

-400
Belgium

Ireland

Korea
UK

Finland

Australia

Norway
Switzerland

Canada
Denmark

Sweden
Austria
France

Portugal

Slovenia
Germany
Greece

Italy

Spain

Czech Republic
Japan

United States

New Zealand
Netherlands

Source: IMF

4.78 Fourth, the exposure of the public finances to wider economic risks is a potential factor. The
Icelandic and Irish governments were running fiscal surpluses and forecasts for gross debt in
2010 were below 30 per cent of GDP for both countries prior to the financial crisis. Yet both
were forced by the high fiscal costs of dealing with it to seek support from the IMF and
European Union.60 This was because of their unusually large banking sectors whose
liabilities were, in effect, a contingent liability of the Government. When the financial crisis
hit, these liabilities were transferred to the Government, which was unable to service them
without outside assistance. Governments whose revenues depend heavily on exports of
volatile or finite resources such as fossil fuels can also find themselves quickly plunged into
debt distress when either prices or volumes fall. The UK does, of course, have a relatively
large financial sector, though it is considerably more resilient today than at the time of the
financial crisis.61

59
Yousefi, S. R., Public Sector Balance Sheet Strength and the Macro Economy, IMF Working Paper, August 2019.
60
IMF, Fiscal Transparency, Accountability, and Risk, August 2012.
61
See Chapter 3 of our 2019 Fiscal risks report.

191 Fiscal risks report

Cost of public debt

4.79 Fifth, government s institutional capacity to deliver large and rapid fiscal adjustments if such
risks crystallise also shape investor confidence. The G
the state of the public finances to respond to fiscal shocks is important to reassure investors
that it can deal with future fiscal pressures without resorting to default or inflationary
measures. The degree of fiscal centralisation/decentralisation both within and across levels
of government, starting levels and buoyancy of the tax burden, and the extent of structural
rigidities in government expenditure all play roles
deliver o
for doing so is discussed in the next section (from paragraph 4.82).

4.80 Sixth, on the demand side of the market, the demand for
government debt may be a factor. Bonds offer a safer way of transferring purchasing power
over time than risky assets, such as equities. They are therefore a natural asset for
institutions with fixed future liabilities, such as defined-benefit pension funds, to hold. But in
addition, many investors want to hold high-quality government bonds because they can be

as the stock of debt grows, providing another reason why yields may increase with the stock
in and of
itself, but demand may be reinforced by regulatory requirements imposed on financial
institutions. Central bank purchases of government bonds under quantitative easing
programmes have also provided an important additional source of demand in recent years.
Looking to international investors, the UK has the additional advantage of being a reserve
currency (albeit a rather junior one) and UK gilts, like US Treasuries, have often benefitted
from being seen as offering a safe haven at times of global stress. Of course, that might no
longer be the case if the UK alone was subject to a shock with major fiscal consequences.

4.81 Last but certainly not least, a credible institutional framework for macroeconomic
policymaking is central to maintaining
a sustainable footing and that monetary policy will deliver low and stable inflation.
Quantitative research has found that institutional strength and transparent fiscal frameworks
are correlated with reduced borrowing costs, for both emerging and advanced economies.62
The UK has historically been a leader in fiscal transparency, as noted by the IMF in their
December 2020 assessment of fiscal space, which judged that the UK benefitted from the
strong macroeconomic and fiscal forecasting capacity a long-
standing and credible medium-term budget framework 63

Primary balance adjustment

4.82
ability and willingness to make significant adjustments to the public finances when
necessary. When the fiscal adjustment required to arrest a growing burden of debt exceeds
what governments have demonstrated they can achieve, taking into account prevailing

62
See Alfonso, A., and Tovar Jalles, J., Fiscal Rules and Government Financing Costs, Fiscal Studies, March 2019 on advanced
economies, and Kemoe, L. and Zhan, Z., Fiscal Transparency, Borrowing Costs, and Foreign Holdings of Sovereign Debt, IMF Working
Paper, August 2018 for a discussion of the effects of fiscal transparency in emerging economies.
63
IMF, United Kingdom: 2020 Article IV Consultation-Press Release; Staff Report; Staff Supplement; and Statement by the Executive
Director for the United Kingdom, December 2020.

Fiscal risks report 192

 Cost of public debt

political and institutional constraints, governments are likely to find it increasingly difficult to
sell their debt to sceptical investors. This section reviews some of the evidence on the size
and speed of past episodes of fiscal adjustment in the UK and other advanced economies.

4.83 Chart 4.18 shows the distribution of year-on-

1700. The vast majority (83 per cent) of these annual changes in the primary deficit are
between minus 2 and plus 2 per cent of GDP, with larger adjustments almost always
associated with wars and other major crises. For instance, the most significant annual fiscal
consolidation of the past century was a fall of 9 per cent of GDP following World War II,
which was the result of moving from the demands of total war to peacetime. Moreover,
these changes in the primary balance represent just a single year of adjustment sustaining
a similar tightening in fiscal policy for several years would be increasingly difficult (as
illustrated by the austerity fatigue that built progressively over the past decade). So, for
instance, even the significant fiscal consolidation following the financial crisis peaked at just
over 2 per cent of GDP year-on-year.

Chart 4.18: UK year-on-year change in the primary surplus

120
Fiscal loosening Fiscal tightening
100
Number of years since 1700

80
Other periods
Crisis periods
60
Post-crisis periods

40

20

0
Less -10 -9 to -8 to -7 to -6 to -5 to -4 to -3 to -2 to -1 to 0 to 1 to 2 to 3 to 4 to 5 to 6 to 7 to 8 to 9 to More
than to -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 than
-10 Year-on-year change in the primary surplus, per cent of GDP +10

Note: Crisis periods include the post-Napoleonic depression, World War One, World War Two, the financial crisis and the
coronavirus pandemic. Post-crisis periods are the three years following each crisis.
Source: Bank of England, ONS, OBR

4.84 A sharper reduction in the deficit is more likely when growth is strong because fiscal drag
makes tax revenues rise faster than GDP, while spending typically falls, for example due to
lower payments of unemployment benefits. Chart 4.19 therefore compares changes in the

18). Predictably, this confirms that an improvement in the

primary balance is much less common in the absence of strong growth (shown by more
data points in the top right quadrant of the chart than in the bottom right). It also shows
that, while the 2000s have already seen two instances of the primary surplus falling by more
than 2 per cent of GDP, it has not yet seen any increases in the primary surplus on that
scale. This underscores the asymmetric nature of fiscal shocks that the UK has faced in
recent decades.

193 Fiscal risks report

Cost of public debt

Chart 4.19: UK year-on-year change in the primary surplus and nominal GDP
growth excluding crisis periods
1800s 1900s 2000s
25
Positive GDP growth, Positive GDP growth,
20
decrease in primary surplus increase in primary surplus
15

10
Nominal GDP growth

-5

-10

-15
Negative GDP growth, Negative GDP growth,
-20
decrease in primary surplus increase in primary surplus
-25
-5 -4 -3 -2 -1 0 1 2 3 4 5
Year-on-year change in the primary surplus, as a proportion of GDP
Note: Excluding crisis periods and the three years following each crisis. Crises excluded are the post-Napoleonic depression, World
War One, World War Two, the financial crisis and the coronavirus pandemic.
Source: Bank of England, ONS, OBR

4.85
where the year-on-year changes in the primary deficit have also been highly concentrated
between 2 and -2 per cent of GDP (Chart 4.20). Again, the transition from the needs of a
wartime economy to peacetime accounts for the most significant episodes of fiscal
tightening in the data, with the United States and Canada seeing improvements in the
primary balance of 15 and 21 per cent of GDP respectively in 1946.

Chart 4.20: G7 year-on-year change in primary deficit

400
Fiscal loosening Fiscal tightening
350
Canada
France
300
Germany
Italy
Number of years

250
Japan
200
UK
US
150

100

50

0
Less -10 -9 to -8 to -7 to -6 to -5 to -4 to -3 to -2 to -1 to 0 to 1 to 2 to 3 to 4 to 5 to 6 to 7 to 8 to 9 to More
than to -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 than
-10 +10
Year-on-year change in the primary surplus, per cent of GDP

Note: Data since 1800 or earliest available date

Source: IMF, OBR

Fiscal risks report 194

 Cost of public debt

4.86 Institutional arrangements for fiscal policymaking also affect the achievability of a significant
G-20
countries with stronger budget institutions overall have tended to plan and deliver more fiscal
adjustment in the wake of the [financial] crisis
institutions delivered, on average, a 2¼ per cent of GDP reduction in the cyclically adjusted
primary balance from 2010 to 2012, compared to the ¼ per cent of GDP reduction
64
Th -standing commitment to
fiscal transparency and medium-term budgetary planning is likely to be an advantage here

experienced a strong record of delivering on its medium-term commitments on the

expenditure side (while also noting that revenue forecasts had tended to be overly
optimistic).65 However, no consolidation is painless and, ultimately, how quickly and how far
the Government can adjust is also determined not only by institutional arrangements but
also by political factors.

Demand for UK government debt

4.87 As also noted above, another factor that may be important during a sovereign debt crisis is

base. This section provides further analysis of the buyers of UK gilts and how the
composition of holders has changed over the past 35 years (Chart 4.21). It also considers
the likely stickiness of that demand in the face of a UK-specific crisis.

4.88
since the financial crisis as of the final quarter of 2020 the Bank held 32 per cent of the
stock of gilts in issue, with a market value of 38 per cent of GDP. Since the start of the
pandemic, the Bank has in effect absorbed 83 per cent of net gilt issuance (the second
largest purchasers have been overseas investors, who have purchased 14 per cent).66 The
G
debt has risen rapidly (see Box 4.5). The Bank estimated that the initial £200 billion tranche
of quantitative easing in 2009 lowered 10-year gilt yields by around 1 percentage point.67
Subsequent tranches appear to have had a somewhat smaller impact, though the Bank
estimates that gilt purchases during the pandemic have had the largest impact on gilt yields
since the financial crisis, lowering them by almost 0.4 percentage points.68

4.89 If and when the Bank decides to run down the APF either by active sales or by allowing it
to run off organically as the gilts mature it can be expected to put upward pressure on
yields (though this may be partially offset by banks wanting to replace the reduced stock of
central bank reserves with other liquid assets such as gilts). There is a risk that the

64
See IMF Policy Paper, Budget Institutions in G-20 Countries An Update, April 2014 for more detail on assessments of institutional
strength.
65
IMF, United Kingdom: Fiscal Transparency Evaluation, November 2016.
66
Calcul published
by the ONS).
67
Joyce, M. Lasaosa, A., Stevens, I and Tong, M., The Financial Market Impact of Quantitative Easing in the United Kingdom, International
Journal of Central Banking, September 2011.
68
Bailey, A., Bridges, J., Harrison, R., Jones, J., and Mankodi, A., The central bank balance sheet as a policy tool: past, present and
future, Staff Working Paper No. 899, December 2020.

195 Fiscal risks report

Cost of public debt

movement in yields when such a policy is announced could be quite sharp, as market

triggered by speculation that the Federal Reserve was about to reduce the pace of its asset
purchases).

Chart 4.21: UK gilt holders

125
Other
Monetary financial institutions
Insurance companies and pension funds
100
Other financial institutions
Overseas holdings
Bank of England
75
Per cent of GDP

50

25

0
1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019
Note: Chart shows holders of gilts at market value. Other includes local government, non-financial corporations, households and
non-profit institutions serving households.
Source: DMO, ONS

4.90 The second biggest development has been the increase in the proportion of debt held by
overseas investors, from 11 per cent in the first quarter of 1987 to 28 per cent by the final
quarter of 2020. While this is a substantial increase, the proportion of UK debt held by
overseas investors is still below the average for advanced economies (Chart 4.22). It is,
though, much closer to the average of advanced economies excluding the euro area (where
cross-holdings are more likely), which is 29 per cent.

Fiscal risks report 196

 Cost of public debt

Chart 4.22: Foreign holdings of general government debt securities

100
Q4 2007
90
Q4 2020
Per cent of general government debt

80
Q4 2020 average
70
Q4 2020 average (excl. euro area)
60

50

40

30

20

10

0
Belgium
Finland

Norway

Switzerland
Ireland

Korea
Australia

Canada
Slovenia

Sweden
France

Germany

United Kingdom
Austria

Denmark

Italy

United States
Portugal

Spain

Japan
Greece

Czech Republic
Netherlands

New Zealand
Source: IMF

4.91 A third notable change has been the rise in domestic bank holdings of gilts from 2 per cent
in 2007, pre-financial crisis, to an average of 6 per cent in 2020. In part that reflects the
Basel 3 banking regulations. Higher
holdings of government debt by domestic banks increase the risk of a government-bank

ing the potential cost of official support, and

in turn worsening the expected fiscal position). The proportion of UK debt held by domestic
banks is, however, still low compared to countries where such adverse feedback effects were
experienced in the past. For example, during the euro area debt crisis, domestic bank
holdings of government bonds in Portugal, Italy, Ireland and Spain peaked between a
quarter and a third of total debt, and a fifth for Greece.

4.92 Fourth and finally, pensions funds have been, and continue to be, reliable holders of gilts,
in part for regulatory reasons. Their holdings have stayed relatively constant at around 25
per cent of GDP over the past 35 years, although that means their share of the total stock in
issue has fallen as the debt-to-GDP ratio has risen. The pensions landscape has been
changing for many years, with defined benefit (DB) schemes in decline and defined
contribution (DC) schemes growing. DC schemes currently invest more in equities and less
in government bonds compared to DB schemes.69 Given the difference in maturity of the
two different types of schemes and regulatory needs for DB schemes to hold gilts, it is likely
that pension funds will remain a stable source of demand for gilts in the future.

69
Pension Policy Institute, DC scheme investment in illiquid and alternative assets, March 2019.

197 Fiscal risks report

Cost of public debt

Chart 4.23: Defined benefit pension schemes asset allocations

Government conventional bonds Government index-linked bonds Equities
Corporate fixed interest Other investments
100

90

80
Per cent of total assets

70

60

50

40

30

20

10

0
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Source: PPF, OBR

4.93 The changing sectoral structure of gilt holders in recent decades has increased some risks
and reduced others. Greater exposure to foreign holdings may have made the UK
somewhat more vulnerable to sudden changes in investor sentiment in a crisis. By contrast,
greater holdings by the Bank of England have reduced risks (by reducing yields and directly
reduced rollover risk and by helping to preserve a deep and liquid market for gilts in times
of crisis). And pension funds and insurance companies remain large and dependable
sources of private domestic demand for gilts. However, the large Bank of England holdings
mean that the fiscal position is more sensitive to variations in Bank Rate. Consequently, were
a loss of investor confidence to necessitate a tightening in monetary policy (for instance to
prevent a sharp fall in the pound), then it would have immediate implications for the fiscal
position (as illustrated in our final scenario).

Simulating a loss of investor confidence in the UK

4.94 Public borrowing and indebtedness have risen substantially as a result of the pandemic. But
the former is likely to drop sharply as the economy rebounds and pandemic support
measures come to an end. And the latter, while markedly higher than before the financial
crisis, respond to the pandemic (see Box 1.1 in
Chapter 1). And while there are certainly fiscal challenges facing the Government, including
dealing with the rising costs of an ageing population (see our past FSRs) and legacy of the
pandemic (see Chapter 2 of this FRR), addressing them does not look unmanageable, as
can be inferred from the continued robust demand from investors to hold UK gilts. So there
seems to be very little immediate danger of the UK being subject to a sovereign debt crisis.

4.95 Nevertheless, in the spirit of considering even quite remote tail risks, our final scenario looks
at what might happen if, for some reason, there were a loss of investor confidence in the UK

Fiscal risks report 198

 Cost of public debt

G creditworthiness and its macroeconomic policy framework. In this scenario,

rising public debt leads to growing concerns amongst bond investors about the risk of
losses. This in turn pushes borrowing costs up further, generating a vicious circle of rising
debt and rising borrowing costs. In such a scenario, the Government would need to
undertake significant fiscal tightening to stabilise its debt-to-GDP ratio. As discussed earlier,
adjustments greater than 2 per cent of GDP are rare outside of the automatic fiscal
correction that occurs after wars end (and as will happen as pandemic support rolls off).

4.96 As in our other scenarios, our aim is to highlight how the fiscal consequences of a loss of
confidence could unfold rather than take a view of the potential cause of the crisis. That
said, such a crisis would be more likely to be triggered by a UK-specific shock rather than a
global one, such as the pandemic. Global shocks typically lead investors to shift from risky
assets into high-quality government debt, benefitting countries perceived as safe havens. In
the case of an idiosyncratic shock affecting just the UK, there is more likely to be a flight of
investors from UK gilts into overseas assets instead.

4.97 In this scenario, we assume:

There is an adverse supply shock that pushes real GDP growth 4 percentage points
below our baseline for two years a similar output loss to the financial crisis and less
than half that precipitated by the pandemic. This results in real GDP growth of 1.0 per
cent in 2022-23 and -2.5 per cent in 2023-24. Rising borrowing costs and an
escalating crisis mean that growth fails to rebound and the economy continues to
shrink for a further two years. We assume growth then returns to our baseline rate but
there is no catch-up growth so long as the crisis continues. Output consequently lies
persistently beneath the baseline, with the shortfall peaking at 8.5 per cent.

Inflation rises to 4 percentage points above baseline in 2022-23, reflecting both the
shock and a depreciation in sterling as investors sell UK assets. Inflation subsequently
falls back to target over three years.70

The Bank of England reacts to the rise in inflation by raising Bank Rate to 4.1 per cent
in 2022-23 (4 percentage points higher than the baseline). It gradually falls back to
baseline by 2027-28 as inflation returns to target.71

In line with the assumptions in our previous scenarios, the primary balance worsens in
the first two years due to falling tax receipts as GDP growth weakens, welfare spending
rises though non-welfare spending is fixed in nominal terms. After 2024-25, we
assume tax receipts remain constant as a share of GDP. We assume primary spending
remains the same as our baseline and so does not fall with GDP. This allows us to
calculate the reduction in primary spending or increase in taxes that would be
necessary to compensate for rising interest costs and stabilise the debt ratio, rather
than the scenario delivering this adjustment by assumption.

70
As modelled, higher inflation delivers a fiscal benefit for the first two years as tax receipts increase with nominal GDP and non-welfare
spending is fixed in nominal terms. Since inflation is due to the depreciation in sterling, the receipts benefit would probably be much
smaller because it would lead to reverse fiscal drag. But such effects would small relative to the overall scenario.
71
Estimated using the model described in Working paper No.4: A small model of the UK economy, OBR, July 2012.

199 Fiscal risks report

Cost of public debt

We assume that private sector bailouts in the first year of the crisis add 10 per cent of
GDP to debt (a stock-flow adjustment ). This is comparable to the global financial
crisis where financial sector interventions totalled 9 per cent of GDP.72

The average gilt rate increases as the debt-to-GDP ratio rises because investors
demand an escalating premium to hold UK government debt. We use the results of a
study by Bayer, Born and Luetticke to calibrate the link between debt and the cost of
borrowing, with a 10 per cent increase in government debt leading to a 250 basis
point increase in the yield on government debt in the short run, fading to a 25 basis
point increase in the long run.73 As a greater reliance on short-dated debt is often
necessary in severe crises such as we model here, we also assume that one third of the
new debt issued by the Government during the crisis has a maturity of one year.74
Issuing short-dated debt means that subsequent increases in interest rates feed
through faster into public spending. Chart 4.24 shows the profile for average gilt rates
in the scenario.

Chart 4.24: Interest rates in the loss of investor confidence scenario

12
Average gilt rate (baseline)
Average gilt rate (scenario)
10
Bank Rate (baseline)
Bank Rate (scenario)
8
Per cent

0
1998-99 2001-02 2004-05 2007-08 2010-11 2013-14 2016-17 2019-20 2022-23 2025-26 2028-29
Note: Average gilt rate is the average yield of gilt issuance before 2012-13
Source: Bank of England, DMO, OBR

4.98 Net interest payments rise rapidly in this scenario, reaching 9.5 per cent of GDP by 2029-
30 higher than in any year in at least three centuries (the previous peak was 8.4 per cent
of GDP in 1926-27). Borrowing therefore increases throughout the scenario due to higher
primary borrowing as the economy shrinks and interest costs escalate (Chart 4.25),
reaching 15 per cent of GDP in 2029-30. The debt-to-GDP ratio also rises in every year of
the scenario and reaches 162 per cent of GDP by 2029-30 (Chart 4.25). While higher debt
burdens were witnessed in 24 years of the twentieth century, the cost of servicing debt is

72
See Annex B of our March 2021 Economic and fiscal outlook, calculated as a percentage of 2009-10 GDP.
73
Bayer, C., Born, B., and Luetticke, R, The Liquidity Channel of Fiscal Policy, CEPR Discussion Paper 14883, 2021.
74
This simple assumption is designed to produce an effect similar to the shortening of maturity that happens when governments need to
issue large amounts of debt.

Fiscal risks report 200

 Cost of public debt

higher than at any point during that period. This demonstrates that it is the cost of servicing
debt, not the debt-to-GDP ratio alone, that is central to generating financing pressures. We
only show the scenario until 2029-30, as the Government would most likely need to
undertake major fiscal retrenchment to stabilise the debt (discussed below) or seek support
from multilateral or bilateral creditors (such as the IMF) long before this point is reached.

4.99 Rising interest costs are primarily the result of the adverse feedback loop between higher
debt and higher gilt rates. The initial shock causes debt to rise 12 per cent of GDP in 2022-
23, mainly due to the private sector bailouts (Chart 4.26).75 This rise in debt pushes up the
average gilt rate, which causes the G
increase further. By 2029-30, the average gilt rate reaches 9.6 per cent (10-year gilt rates
were last at this level in 1991). The shortening of the maturity of gilts means higher market
rates feed through into higher public spending more quickly. In the first few years of the
scenario, higher inflation and the rise in Bank Rate also contribute to the sharp increase in
interest costs.

Chart 4.25: Loss of investor confidence scenario: key outputs

10 10
Effective interest rate less nominal GDP Net interest payments
8 growth (R-G) 9

8
6
7
4
Per cent of GDP

6
2
Per cent

5
0
4
-2
3
-4 2
-6 1
-8 0
2000-01 2005-06 2010-11 2015-16 2020-21 2027-28 2000-01 2005-06 2010-11 2015-16 2020-21 2027-28
20 180
PSNB PSND
160
15 Loss of confidence scenario
140
Baseline
Per cent of GDP
Per cent of GDP

120
10
100

5 80

60
0
40

-5 20
2000-01 2005-06 2010-11 2015-16 2020-21 2027-28 2000-01 2005-06 2010-11 2015-16 2020-21 2027-28
Source: ONS, OBR

75
In 2022-23, real GDP falls 4 per cent below baseline and inflation rises to 4 per cent above baseline, meaning nominal GDP is
unchanged. We use changes in nominal GDP to calculate changes in the primary deficit, which means the primary deficit is the same as
the baseline in this year.

201 Fiscal risks report

Cost of public debt

Chart 4.26: Loss of investor confidence scenario: differences from baseline

14 90
PSNB PSND
80

Per cent of GDP difference from baseline

12
Per cent of GDP difference from baseline

Primary deficit Stock-flow adjustments

70
Net interest payments Growth-corrected interest rate (R-G)
10
PSNB 60
Primary deficit
8 50
PSND
6 40

30
4
20
2
10

0 0
2022-23 2024-25 2026-27 2028-29 2022-23 2024-25 2026-27 2028-29
Source: OBR

4.100 In this scenario, it is possible that liquidity in the gilt market could dry up as investors
struggle to price gilts accurately. March 2020 saw such incipient illiquidity, leading the Bank
of England .76 Our scenario ignores such
illiquidity problems, implicitly assuming that the Bank would again step in if needed. But
were that not to happen, it could lead to a as buyers leave the market
altogether.

4.101 As noted at the start of this section, the Government would need to make large primary
deficit adjustments to stabilise debt in this scenario. Assuming this would only happen after
the economy stops shrinking in 2026-27, Chart 4.27 shows that the necessary adjustments
would have to be very large by historical standards. The 5 per cent of GDP reduction in the
primary deficit needed to stabilise debt in 2026-27 has only occurred in 3 per cent of the
years since 1700 and all of these were after wars. The required adjustment increases to 8
per cent in 2029-20 as higher interest costs have to be offset by an even larger primary
surplus to stabilise debt. So acting earlier both reduces the eventual adjustment needed and
stabilises debt at a lower level. It may not in practice be necessary to make a full adjustment
in any individual year to break the feedback loop between rising debt and rising rates.
Instead, credible plans spread over several years may be sufficient to restore lost investor
confidence, allowing interest rates to fall back and making the required adjustment smaller.

76
For an account of events see Hauser, A., Seven Moments in Spring: Covid-
sheet operations, Bank of England Speech, 4 June 2020.

Fiscal risks report 202

 Cost of public debt

Chart 4.27: Primary surplus adjustment required to stabilise the debt-to-GDP ratio
9

6
Per cent of GDP

0
2026-27 2027-28 2028-29 2029-30
Note: Chart shows the change in the primary surplus required to stabilise PSND excluding the Bank of England to avoid distortions
from the rundown of the Term Funding Scheme.
Source: OBR

Box 4.5: The fiscal impact of the Asset Purchase Facility

The Asset Purchase Facility (APF) houses the assets (predominantly gilts) purchased by the Bank
of England as part of its programme of quantitative easing initiated in 2009. QE has had a
variety of impacts on the public finances, for example: indirectly through its impact on the wider
economy; through lowering yields on government debt; and through shortening the effective
maturity of . This shortening of maturity arises as Bank
reserves (floating rate debt) have been used to purchase gilts (long-term fixed-rated debt). This
has increased the short-run sensitivity of debt interest payments to rate rises. To date, the APF
has benefitted the Treasury (which receives the cash surplus) as the higher rates on longer-dated
debt have meant that the payments on the additional reserves have fallen far short of the
payments on the associated gilts. Our March forecast shows the APF paying £0.6 billion in
2022-23 on the £875 billion reserves issued by then to finance gilt purchases (a rate of under
0.1 per cent), but receiving £17.2 billion on the purchased gilts (a yield of 2.0 per cent). This
reduces overall public sector interest payments to the private sector (and therefore the deficit) by
£16.6 billion. To date, positive cash flows from the APF to the Treasury have totalled £113
billion.
It should be noted, though, that despite the large direct reduction in debt interest costs associated
with the APF (reducing the effective interest rate by about 0.8 percentage points), the total fall in
debt interest costs over the past decade is much larger (4.1 percentage points). The overall fall
reflects lower Bank Rate and a flatter yield curve, in part a consequence of quantitative easing.
In the scenarios used in this chapter, we hold the size of the APF constant. Changes to Bank Rate

203 Fiscal risks report

Cost of public debt

receipts and payments. In the baseline, net interest savings gradually decline, at first largely
because Bank Rate rises and then as gilts mature and are rolled over at lower rates (Chart D).
In the higher R and higher R and G scenarios, a rising Bank Rate sharply increases payments
on reserves reducing the cash surplus of the APF and by 2026-27 these payments exceed the
coupon income earned and so the APF shows a deficit. Gradually, as gilts are rolled over at

pattern, though here, gilt rates rise far enough that eventually the APF returns to surplus.

Under the scenario, sharp increases in Bank Rate quickly send the
APF into deficit. But as the rise in Bank Rate is only temporary, after a few years the APF returns
to surplus. At the start, scenario is similar, but here the surplus
keeps rising as the soaring gilt rate rapidly increases earnings on rolled over gilts.
Chart D: Net savings to the public sector of the APF in our scenarios
35
Net interest saving to the public sector due to APF
25 Baseline
Higher R and G and Higher R scenarios
15
Temporary inflation shock scenario
Persistently higher inflation scenario
Loss of investor confidence scenario
5
£ billion

-5

-15

Net interest cost to the public sector due to APF

-25
2021-22 2025-26 2029-30 2033-34 2037-38 2041-42 2045-46 2049-50
Note: Loss of investor confidence scenario only shown to 2029-30
Source: OBR

In practice, were inflation pressures to pick up markedly, the MPC might choose to reduce the
size of the APF rather than relying solely on Bank Rate to tighten monetary policy. This would be
consistent intended not to reduce the stock of purchased
assets until Bank Rate reached around 1.5% ,a though the Governor has noted that this guidance
is currently under review. This would have several effects on the APF:
Bank Rate would need to rise less in order to meet the inflation target, resulting in a
larger surplus/smaller deficit.
The smaller size of the APF would correspondingly reduce the size of future surpluses or
deficits.
Gilt sales would be likely to take place at a lower price than was originally paid, leading
to a cash loss (on the assumption that sales only take place once Bank Rate has exceeded

Fiscal risks report 204

 Cost of public debt

1.5 per cent)

good by the Treasury. Such trading losses could potentially be large: as of April 2021,
selling all gilts in the APF portfolio at par would crystallise a trading loss of £114 billion.
The extent to which APF sales could substitute for increases in Bank Rate will depend on how fast
sales could take place. In practice a run-down is likely to take several years, as the MPC has
indicated that any reduction in the stock of assets will be at a gradual and predictable pace .b
In principle, the Government could cover any reduction in the cash flow from the APF by raising
taxes, reducing spending, or increasing borrowing. Commentators have also advanced several
alternative suggestions for mitigating the reduction in cash flow itself:
Lowering (possibly to zero) the interest paid on reserves.c This is economically equivalent
to maintaining the existing arrangements and introducing a new tax on banks related to
their reserve holdings. Also, because it would increase the opportunity cost of holding
reserves relative to other assets, it would change the nature of the monetary transmission
mechanism and force the Bank of England to change its technique of managing short-
term interest rates. It would also divert financial flows from banks to other less transparent
channels.
Require banks to hold some minimum level of reserves but pay a low or zero interest rate
on them, paying Bank Rate on only the excess.d This would maintain the effectiveness of
Bank Rate as a monetary policy tool, b
profitability of acting like a tax on reserve holdings.
Delaying the fiscal consequences of higher interest rates, for example by exchanging
some of the reserves for short-maturity gilts.e This would reduce the speed with which
interest rate changes fed through to spending, but would come with increased funding
risk, since reserves do not need to be rolled over whereas the new securities would.
In conclusion, a tightening in monetary policy is likely to result, one way or the other, in a
smaller contribution of the APF to the public finances. And though there are ways this could be
mitigated, they each involve drawbacks of their own.

a
Monetary policy summary and minutes of the monetary policy committee meeting ending on 20 June 2018, Bank of England.
b
Monetary policy summary and minutes of the monetary policy committee meeting ending on 20 June 2018, Bank of England.
c
Goodhart, C., Evidence to House of Lords Select Committee on Economic Affairs, 16th March 2021.
d

Designing a New Fiscal Framework: Understanding and Confronting Uncertainty, (edited by J.

Chadha, H. Küçük and A. Pabst), 2021.
e
Allen, W., Managing the fiscal risk of higher interest rates, NIESR policy paper 25, 2021.

Conclusion
4.102 Interest rates on advanced economy government bonds have declined for several decades.
Since the mid-1990s, this has been almost entirely attributable to a fall in real interest rates
rather than lower inflation. As interest rates have declined faster than economic growth, this
has also moved the growth-corrected interest rate well into negative territory. This has been

205 Fiscal risks report

Cost of public debt

true in the UK too, where the decline in interest rates has resulted in a fall in debt interest
costs despite the rise in debt from the financial crisis and the pandemic.

4.103 Several explanations have been put forward for this fall in bond rates, including
demographics, the productivity slowdown and shifts in portfolio preferences, but there is still
disagreement about their relative importance. This uncertainty carries over into uncertainty
about the future path of rates. Some factors that affect the savings-investment balance,
particularly demographics, may be starting to reverse though the impact on rates is likely to
be felt only gradually at best. But other forces, especially those affecting portfolio
preferences, could reverse more rapidly.

4.104 Our first two scenarios explored the fiscal consequences of a gradual rebound in real
interest rates. Higher interest rates on their own (say because of a shift in portfolio
preferences) would add to the fiscal headwinds facing the Government. But a rise in interest
rates that is associated with faster growth (perhaps driven by a revival in productivity growth)
would produce a more benign outcome; although debt interest costs rise, that is offset by
the faster expansion in the size of the economy, so that the path of the debt-to-GDP ratio is
a little lower, although still above its pre-pandemic level after 30 years.

4.105 Our third and fourth scenarios explored the fiscal consequences of a rise in interest rates
that is instead associated with higher inflation. Temporarily higher inflation produces a
small reduction in the debt burden, though a good part of that arises from the short-run
cash limits on government spending. Persistently higher inflation produces a similar
outcome in the short to medium run, but is actually counterproductive in the long run
because of an assumed rise in the inflation risk premium on nominally-denominated bonds.
In both scenarios the share of revenues consumed by interest payments rises. The relative
ineffectiveness of inflation in reducing the debt burden reflects several factors, including the

quantitative easing and the relatively high share of indexed-linked bonds.

4.106 Finally, we explored the consequences of a loss of investor confidence leading to a debt
crisis. Currently the demand for UK gilts is fairly robust and the risk of a debt crisis in the UK
in the near future seems remote. But debt crises do happen from time to time, even in
advanced economies, so in line with the emphasis of this Fiscal risks report on catastrophic
risks, we also modelled a scenario in which investor confidence is lost, leaving debt interest
costs and the debt-to-GDP ratio spiralling higher. In these conditions an unsustainable
position can develop quickly, and early action to halt the spiral is desirable. But, in our
admittedly extreme scenario, the size of fiscal adjustment necessary exceeds that which has
been achieved in the past century (outside of the large automatic corrections that occur after
the end of major wars and also expected to be the case as the pandemic subsides and
support measures such as the CJRS are wound down). That speaks to the importance of
avoiding triggering such a spiral in the first place, by maintaining investor confidence in
the G to monetary and fiscal responsibility and the institutions that
support them.

Fiscal risks report 206

5 Update on other fiscal risks

Introduction and summary

5.1 The preceding chapters of this report have focused on three large and looming threats to
the public finances, but the array of other fiscal risks highlighted in previous reports have
not evaporated. Indeed, as this chapter shows and as history warns, the pandemic has
triggered the crystallisation of several of these risks, aggravated many others, and even
diminished a few. This final chapter details how these other risks have evolved since our
previous Fiscal risks report (FRR) in July 2019 and how our full risk assessment has changed
after factoring in both those changes, and the risks discussed in the preceding chapters.

5.2 Our 2019 report was accompanied by our first risk register, which identified 106 risks from
the 57 issues we raised in our 2017 FRR plus additional ones from the 2019 report. The
Charter for Budget Responsibility requires the Treasury to respond formally to the FRR within
2018 Managing fiscal risks report was a substantive
140-page response to our initial report. The response to our 2019 FRR was
understandably overtaken developing and
delivering the economic policy response to the crystallisation of what has proved to be the
largest fiscal risk in peacetime. Its official response to the 2019 FRR therefore constituted a
brief Written Ministerial Statement by the Chancellor on 14 July 2020, which discussed only
four issues that in one way or another related to 12 of the 106 risks we had identified.1

Changes in fiscal risks since 2019

5.3 For this report, we have first recast and consolidated some of the risks that were identified
on our 2019 register, bringing the total down to 97. Of these:

14 have crystallised including weaker productivity growth, lower net migration, and the
declining proportion of spending subject to firm DEL controls. Of these, 13 remain
active risks in future (including normal cyclical downturns, the deterioration in public
sector net worth, and cost overruns for major projects) and 1 has been removed (the
balance sheet risk relating to the classification of housing associations).

19 have increased, including those related to higher future health and social care
spending as a result of the pandemic, the longer-term sustainability of the fuel duty tax
base in light of the bringing forward of the ban on petrol-driven cars, and the
pandemic-driven increase in the non-payment of taxes due.

1
The issues that were referenced were weak productivity (which accounted for two of the 90 risks), climate change (six risks), shadow
banking (three risks) and tax reliefs (one risk).

207 Fiscal risks report

Update on other fiscal risks

11 have decreased, including the tendency for fiscal policy to respond asymmetrically
to movements in our underlying forecasts following the tax rises announced in the
March Budget, the risks associated with persistent household financial deficits in light
of the savings accumulated by some during the pandemic, and the loss of revenue
from people moving to more lightly taxed forms of employment status.

29 remain unchanged, including our broad assessment around risks associated with
the financial sector, which has so far weathered the coronavirus storm, clean-up costs
for nuclear plants, and those around stated policy aspirations.

3 have been resolved and removed from the register, including those around the

for commercial property purchases.

21 have been removed for other reasons including their being unquantifiable,
superseded by analysis presented in this report, or consolidated with other risks (taking
the total number of risks removed from the register to 25).

5.4 Finally, 15 risks have been added in this report including nine arising from the coronavirus
pandemic, three associated with climate change, two relating to the cost of public debt and
a final one on the threat posed by potential cyberattack. This takes the total number of risks
in our register to 87. Chart 5.1 depicts these changes as well as the number of risks that
have been affected to some extent by the pandemic.2

Chart 5.1: OBR fiscal risk register: changes since our 2019 report
35

30

25
Number of risks

20

15

10

0
Crystallised Increased Unchanged Decreased Resolved Removed* Added
(ongoing)
Note: Darker shaded portions show the number of risks within each category that have been affected to some extent by the pandemic.
* The one risk that has crystallised and is no longer on the register is included in 'Removed'.
Source: OBR

2
Chart 5.1 classifies all those risks with a coronavirus impact had an impact.

Fiscal risks report 208

 Update on other fiscal risks

Impact of coronavirus on other fiscal risks

5.5 Reflecting the correlated nature of fiscal risks, of the 97 risks from 2019, 38 have been
affected to some extent by the coronavirus pandemic. This includes around half of the
economy risks, two-thirds of the public spending risks, half of the risks relating to the
and one-third of revenue risks. Of those 38, in 15 cases the
risk has increased and in ten cases it has crystallised (with the remaining 13 being either
unchanged, decreased resolved or removed).

5.6 Examples of those affected, each of which is discussed in more detail below, include:

Pandemic-related pressures on health spending that are described in Chapter 2. These

could amount to £7 billion a year on average over the next three years, with pressures
likely to be greatest in the near term. The larger sources of potential pressure include:
maintaining a standing capacity for test and trace and vaccinations; addressing the
backlog of elective treatments built up during the pandemic; and the implications for
NHS productivity of building in greater resilience and a greater capacity for infection
control than was allowed for in pre-pandemic plans

The state pension triple lock, where unusual pandemic-related fluctuations in earnings
growth have seen it rise to 5.6 per cent in the three months to April 2021, from where
it is almost certain to rise further in the three months before the uprating is calculated.
The triple lock raises spending by £0.9 billion for every 1 percentage point, and our
March forecast assumed uprating of 4.6 per cent next year. So, if earnings growth in
the three months to July period that determines triple lock uprating for next April was 8
per cent, as some expect, that would add around £3 billion a year to spending.

The risk of lower net migration, which has crystallised as the pandemic and associated
lockdown has led to significantly fewer net arrivals into the UK than we previously
expected, with initial modelling suggesting that there was a net outflow of around
67,000 between March and June 2020 alone.

Non-payment of taxes due and the tax gap, where the lockdowns led to sharp rises in
unauthorised tax debt, which were then overlaid by Government support measures
allowing taxpayers to defer self-assessed income tax and VAT payments. To give a
sense of scale, around £34 billion of VAT was deferred between March and June
2020, and just under half had already been repaid by the end of April 2021.

Risks associated with statistical reclassifications, such as the bringing of some train
operating companies into the public sector during the pandemic. The scale of the
impact of this on the government balance sheet has not yet been quantified.3

3
As noted in ONS, Recent and upcoming changes to public sector finance statistics, May 2021, the ONS has so far partially implemented
the reclassification of train operating companies and the full balance sheet impact is not yet known.

209 Fiscal risks report

Update on other fiscal risks

Structure of the chapter

5.7 The remainder of this chapter summarises our updated assessments for each of the 97 risks
from the 2019 risk register and describes the reasons for material changes.4 Risks are
grouped into those deriving from the economy, financial sector, government revenue,
government spending, the government balance sheet, and the fiscal policymaking process.
We also discuss the 21 risks that have been removed and the 15 that have been added in
this report (some that are genuinely new and some that we have assessed for the first time).

Economy risks
5.8 Since our 2019 FRR, of the 13 economy risks that we identified: five have crystallised (of
which four reflect the same risk crystallising over both the medium and long term), two have
increased, one has decreased, one has been resolved, one has been consolidated into
another, and the remaining three are unchanged. Only weak productivity growth was

4
The tables in this chapter present each of the risks from the 2019 risk register and our latest assessment of those risks in terms of the
probability of the risk crystallising and its impact on public sector net debt. The risks are grouped into those that affect the medium term,
within our typical 5-year forecast horizon, and longer-term risks that lie beyond that horizon. Some risks span both the medium and long
term. The note beneath each table briefly explains the methodology, and more information is available in our online fiscal risk register.

Fiscal risks report 210

 Update on other fiscal risks

Table 5.1: Latest assessment of the economy risks identified in our 2019 report
Probability of Impact on Coronavirus Treasury Change
crystallising PSND impact response in risk
Medium term
Weak productivity growth* Medium Medium Maybe ⏺ ◆
Low migration * Medium Medium Yes ◆
Recession* Medium High No ◆
Composition of GDP Medium Low No ⚊
Housing sector exposure Medium Not quantified Yes 🡅
Household financial deficits Medium Not quantified Yes 🡇
Current account deficits Medium Not quantified No ⚊
Output gap mismeasurement High Medium Yes 🡅
Risks from 'no deal' Brexit N/A N/A No 🞫
Long term
Weak productivity growth* Medium Medium Maybe ⏺ ◆
Low migration Medium Medium No ⏺ ⚊
Recession* High Medium No ◆
Note: Refer to the risk register on our website for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium-term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long-term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.
Treasury response refers to the Chancellor's written statement of 14 July 2020. A blank implies it was not discussed.
Coronavirus impact asks whether the pandemic has materially changed our future assessment of the risk.
*Risk remains active despite crystallising.

◆ Crystallised 🡅 Increased ⚊ Unchanged 🡇 Decreased 🞫Removed

5.9 The main changes are:

The fiscal risks associated with weak productivity growth have crystallised in both the
medium and long term. In our March 2020 Economic and fiscal outlook (EFO), we
lowered our long-run productivity growth assumption from 2.0 per cent to 1.5 per cent
a year, having reviewed historical and international evidence.5 We also revised down
our central forecast for the level of productivity due to the scarring effects of the
pandemic (while leaving our long-run productivity growth assumption unchanged). But
while the productivity risk has crystallised since 2019, it remains possible that it could
deteriorate further, for example if the pandemic weighs on productivity growth, the
impact of Brexit is greater than we have assumed, or the post-financial crisis period of
sluggish growth continues, so we consider this to remain a live risk.

The one-in-two chance of a recession in any five-year period has also crystallised,
though the pandemic , with the largest
annual fall in output since 1709. This is another risk that remains active despite
crystallising, as the likelihood of future shocks has not diminished. Indeed, as we

5
See Annex B, Long-term economic determinants, in our March 2020 Economic and fiscal outlook.

211 Fiscal risks report

Update on other fiscal risks

discuss in Chapter 1, the world may in fact be becoming riskier than the historical
experience that underpins the probability and impact reported in Table 5.1.

The risk of lower net migration has crystallised as the pandemic and associated
lockdown has led to significantly fewer net arrivals into the UK than we previously
expected (with both outflows likely to have been higher and inflows lower).
Experimental estimates from the ONS suggest that net migration was negative at
67,000 during the second quarter of 20206 compared with positive net inflows of
271,000 in 2019 as a whole.7 Medium-term risks have increased due to uncertainties
around the economic outlook in the UK and source countries for inward migrants,
including the possibility of continuing travel restrictions, as well as the fact that any
post-pandemic catch-up migration will need to take place under the new, post-Brexit
migration regime, which is more restrictive on average than the previous regime.8

The conclusion of the UK-EU Trade and Cooperation Agreement on 24 December

2020 . We previously
estimated that were the UK to default to trading with the EU under World Trade
Organization rules when the transition period ended on 31 December 2020, then real
GDP would have been 2 per cent lower in 2021 (mainly due to temporary short-term
disruptions). The weakening of productivity over the medium and long term would
have reduced output by around 1½ per cent after five years, rising to 2 per cent in
steady state. The fiscal implications of this would have been to raise borrowing by an
average of 0.5 per cent of GDP between 2020-21 to 2025-26.9

5.10 The other changes to our assessments are:

The direct fiscal exposure to the housing sector has increased following
the announcement of a new mortgage guarantee scheme that was introduced in the
March Budget and is due to run until December 2022.

The uncertainty around real-time output gap estimates and its policy implications has
increased. Public health restrictions to control the pandemic have acted to restrict both
supply and demand, while government support measures have made it harder to
disentangle one from the other. The risk has been amplified by large swings in the
data and the wider difficulty in measuring the economy at present.

The risks associated with persistent household financial deficits have improved.
Households reduced consumption during the pandemic, lowering their outlays, while
some have benefited from generous fiscal support measures protecting their incomes.
The net effect has been to boost household financial wealth. For example, deposits
increased by 14 per cent (£210 billion) between February 2020 and April 2021, while
consumer credit fell 18 per cent (£24 billion) over the same period.

6
ONS, Using statistical modelling to estimate UK international migration, April 2021.
7
ONS, Provisional long-term intentional migration estimates, August 2020.
8
See Box 2.4 of our March 2020 Economic and fiscal outlook.
9
See Annex B, Brexit scenarios, in our November 2020 Economic and fiscal outlook.

Fiscal risks report 212

 Update on other fiscal risks

Financial sector risks

5.11 Our assessment of the fiscal risks associated with the financial sector is unchanged.

Table 5.2: Latest assessment of the financial sector risks identified in our 2019 report
Probability of Impact on Coronavirus Treasury Change
crystallising PSND impact response in risk
Medium term
Financial crises impact Low High No ⚊
Long term #N/A
Financial crises impact Very high Medium No ⚊
Post-crisis regulation loosened over time Low Not quantified No ⚊
Large and concentrated banking system Low Not quantified No ⚊
Shadow banking Very low Low No ⏺ ⚊
Regulation risks Very low Low No ⚊
Note: Refer to the risk register on our website for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium-term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long-term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.
Treasury response refers to the Chancellor's written statement of 14 July 2020. A blank implies it was not discussed.
Coronavirus impact asks whether the pandemic has materially changed our future assessment of the risk.
*Risk remains active despite crystallising.

◆ Crystallised 🡅 Increased ⚊ Unchanged 🡇 Decreased 🞫Removed

5.12 The lack of pandemic-related deterioration reflects two factors:

first, reforms that took place during the post-financial crisis decade meant that the
banking system entered the pandemic much better capitalised, and therefore more
resilient to shocks, than in the 2000s; and

second, the authorities stepped in aggressively to ensure markets continued to function

o support
households and businesses, thereby reducing credit risks around b
(thanks in particular to the £75 billion of government guaranteed loan schemes).

5.13 Financial Stability

Report from December 2020, which argues that the performance of the financial system
through the pandemic reflects the resilience that has been built up since the global financial
crisis, and the extraordinary policy responses of the UK authorities .10

5.14 Of course, reducing fiscal risks that might crystallise via the financial sector by taking on
direct exposure to borrowers via loan guarantee schemes does not reduce the overall fiscal
risks associated with bankruptcies and loan defaults, which will rise to the extent that the
associated losses are no longer borne by lenders and are instead transferred to the state.

10
Bank of England, Financial Stability Report, December 2020.

213 Fiscal risks report

Update on other fiscal risks

Revenue risks
5.15 Of the 17 revenue risks from our previous report, none have crystallised and few have been
affected by the pandemic. Three risks have increased, five have decreased, and nine are
unchanged.

Table 5.3: Latest assessment of the revenue risks identified in our 2019 report
Probability of Impact on Coronavirus Treasury Change
crystallising PSND impact response in risk
Medium term
Income tax reliance on high earners Low Low Maybe ⚊
Stamp duty reliance on top end Low Low Maybe ⚊
Self-employment and incorporations Low Low Maybe 🡇
Excise duties: behaviour or technology change Medium Low No 🡅
Policy non-implementation Very high Low No ⚊
Policy aspirations not yet costed High Medium No ⚊
Reliance on anti-avoidance measures High Low No 🡇
Complexity of tax legislation Low Low No ⚊
Non-payment of taxes due and the tax gap Medium Low Yes 🡅
Tax reliefs: costs continue to rise Medium Medium No ⏺ 🡇
Digitalisation: tax policy challenges Medium Low No ⚊
Digitalisation: administration gains Medium Low No ⚊
Financial services: Brexit impact on tax receipts Very high Low No 🡇
Long term
Self-employment and incorporations Medium Low Maybe 🡇
Tobacco: downard consumption continues High Low No ⚊
Fuel duty: further efficiency improvements High Low No 🡅
Oil and gas decommissioning costs Very high Low No ⚊
Note: Refer to the risk register on our website for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium-term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long-term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.
Treasury response refers to the Chancellor's written statement of 14 July 2020. A blank implies it was not discussed.
Coronavirus impact asks whether the pandemic has materially changed our future assessment of the risk.
*Risk remains active despite crystallising.

◆ Crystallised 🡅 Increased ⚊ Unchanged 🡇 Decreased 🞫Removed

5.16 The five improvements relate to:

The loss of revenue as people move to more lightly taxed forms of employment status.
Policy decisions have reduced the medium- and long-term incentive for individuals to
incorporate and benefit from paying the lower rate of tax on corporate profits (and
dividends) compared to the higher rates of income tax and National Insurance
contributions paid on employment income. The most significant change is the reversal
in the decade-long reduction in the main rate of corporation tax first by maintaining

Fiscal risks report 214

 Update on other fiscal risks

the 19 per cent rate at Budget 2020 (rather than implementing the planned cut to 17
per cent) and then by announcing an increase to 25 per cent from April 2023 in the
March 2021 Budget, the latter raising £17.2 billion a year by 2025-26. Reforms to off-
payroll working (announced at Budget 2018 but implemented this April) also reduce
the incentive. Another factor that might reduce the future attractiveness of
incorporating is that owner-managers have been less generously supported than either
employees or the self-
The reintroduction of a small profits rate tempers the effect of these reforms for those
with profits of up to £250,000 (particularly those at £50,000 and lower that will still
pay a 19 per cent rate).11

Reliance on anti-avoidance and compliance measures with relatively uncertain

costings. While the Government has sought to continue to raise revenue from HMRC
anti-avoidance and compliance measures, it has reduced its reliance on these types of
measures. Anti-avoidance and compliance measures announced in the past two
Budgets are estimated to raise less than £2 billion a year by 2025-26. This is dwarfed
by the close-to £33 billion that is expected to be raised in the same year from the
increases to the corporation tax rate and the freezing of income tax thresholds, the
costings for which are considerably less uncertain than those for typical anti-avoidance
and compliance measures.12 By comparison, in the five fiscal events between Budget
2016 and Budget 2018 the expected yield from anti-avoidance and compliance
measures accounted for around 40 per cent of the total from all revenue-raising
measures across the same period a far higher proportion.13

The high and rising cost of tax reliefs and expenditures, and the poor understanding of
changes over time. At Budget 2020 the Government announced two significant policy
changes: the removal of entitlement to use red diesel and rebated biofuels from most
sectors from April 2022; and a reduction in the lifetime allowance for the business
from March
2020. The combined savings from these two measures rises to £3.5 billion a year by
2024-25. HMRC has also expanded the number of reliefs for which it publishes costs.
That said, the cost of R&D tax credits continues to rise quickly, and there remains an
ongoing challenge around new reliefs, including those to be introduced as part of the
discussed below.

Potential effects of Brexit on the financial sector and the tax receipts it generates. The
fiscal risks from a no-deal Brexit have been averted, so while there remains uncertainty
over the future relationship between the UK and the EU, particularly with regard to
financial services, our assessment is that Brexit-related revenue risks have decreased.

11
For a fuller discussion, see paragraphs A.10 to A.14 in Annex A, Policy measures, of our March 2021 Economic and fiscal outlook.
12
This includes the March 2020 Budget decision to maintain the corporation tax rate at 19 per cent, rather than the planned reduction to
17 per cent.
13
Based on the aggregated costings across all years of the forecast at the time.

215 Fiscal risks report

Update on other fiscal risks

5.17 The three revenue risks to have worsened since 2019 are:

The pressure on excise duty tax bases from behavioural and technological change, in
particular the long-term downward trends in fuel and tobacco consumption. The main
change relates to the former, where the Government has announced that it will bring
forward the ban on sales of new petrol and diesel cars and vans to 2030, ten years
earlier than was mooted in 2019, which affects both the medium and long term. The
pace at which electric car sales have been rising has repeatedly exceeded our forecasts
(as discussed in Box 3.2). Fuel and vehicle excise duties are forecast to raise just under
£39 billion in 2025-26, so the risk is fiscally material particularly over the long term.

Non-payment of taxes due and the tax gap. The lockdown in spring 2020 led to a
sharp rise in tax debt, particularly for PAYE income tax. This was then overlaid by
Government support measures allowing taxpayers to defer self-assessed income tax
and VAT payments. As we reported in our November and March EFOs, much of the
deferred and unpaid tax was swiftly repaid no doubt aided
suite of financial support measures but our forecast does assume that some will
ultimately go unpaid.14 To give a sense of scale, around £34 billion of VAT was
deferred between March and June 2020, and just under half had already been repaid
by the end of April 2021. We also assume a relatively modest medium-term impact on
business rates, but there remains the risk that some businesses may be unable to pay
once the payment holiday ends in March 2022, for example due to the pandemic-
induced jump in online retailing. A second factor contributing to the worsening of this
tax gap risk since 2019 is the continuing uncertainty around elements of the post-Brexit
trading regime. In our November 2020 EFO we included a non-compliance loss of
£0.7 billion in VAT and excise duties in 2021-22 to account for risks around the
operation of the UK border, including the decisions to phase in customs controls and
introduce postponed accounting for most import VAT.15 The eventual outcome of
negotiations around the operation of the Northern Ireland protocol is another
unknown, and one that we do not yet have sufficient information about to quantify for
our forecasts.

5.18 The risk around policy aspirations is ever-present, with new ambitions continuously
replacing or augmenting existing ones. Many of the risks that we outlined in 2019 have now
crystallised the 2 per cent stamp duty land tax surcharge for residential property purchases
by non-UK residents will raise £0.1 billion in 2025-26 and the plastic packaging tax a
further £0.2 billion in the same year. Several Brexit-related policies have also now been
confirmed. In their place are manifesto commitments to raise the National Insurance
threshold, aspirations related to reducing emissions to net zero by 2050, and more.16 Until
these are confirmed government policy we are prevented by legislation from costing them,
though in most cases there is simply insufficient policy detail to arrive at a reasonable and
central estimate. Some of the potentially larger new policy risks include:

14
Proportions vary by tax, with the most fiscally significant being the 7 per cent non-payment rate assumed for PAYE tax debt.
15
This is described in detail in Annex A, Policy measures, in our November 2020 Economic and fiscal outlook.
16
See the Policy risks database on our website for the complete current list.

Fiscal risks report 216

 Update on other fiscal risks

The G7 Finance Ministers agreement around the taxation of multinational company

profits. Pillar One of the agreement grants taxation rights to market countries and
applies to multinationals with global profit margins in excess of 10 per cent. At least
20 per cent of profits beyond the 10 per cent margin will be subject to tax in the
market jurisdiction. Pillar Two is a commitment to a global minimum tax rate of at
least 15 per cent, on a country-by-country basis. The G7 agreement is not legally
binding on non-member countries and the next stage in the process takes place in the
forthcoming G20 meeting. The Chancellor has committed to removing the digital
services tax (DST, forecast to raise £0.4 billion this year, rising to £0.7 billion by 2025-
26) once a Pillar One solution is in place. The Government expects the agreement to
raise revenue, but there remain too many uncertainties to determine a reasonable and
central estimate at this stage, with initial external estimates varying considerably.17

Bank surcharge. ongoing review confirms its initial view that

levying the existing bank surcharge rate on top of the planned 6 percentage point rise
in the corporation tax rate would result in a combined level of bank taxation [that]
would be too high , ensuing policy changes would reduce revenue. We forecast the
bank surcharge will raise £1.4 billion in 2025-26 at a rate of 8 per cent, so roughly
£0.2 billion per percentage point of the existing tax rate.

The Chancellor nouncement on the eight English locations that are

later this year. Freeports are designated areas within a
country that usually lie outside its customs territory. The freeports will benefit from
simplified customs arrangements and duty suspension in approved customs sites, plus
several tax concessions that are yet to be costed. These could include enhanced capital
allowances, and reliefs for stamp duty land tax, employer NICs, and business rates.
Once sufficient detail on these concessions is available, we will be able to factor their
cost into our forecasts.

Uncertainty over the implementation and operation of the Northern Ireland protocol.
The UK-EU Joint Committee that is tasked with overseeing the implementation of the
protocol published operational decisions last December, including several temporary
for the collection of customs duties as requested by the UK
Government. Since then, the UK Government has unilaterally extended some grace
periods, leading the European Commission to express strong concerns over
the action. Current statements from both parties suggest it may be a while before final
implementation, and the ensuing fiscal consequences, are determined.

17
Law firm Clifford Chance estimated that
minimum tax rate of 15 per cent would depend on the exact rules but were likely to fall in a range of £900m to £5bn a year based on 2019
but also that the upper estimate was But the think tank Taxwatch has suggested that
company that is subject to the DST, the Pil .

217 Fiscal risks report

Update on other fiscal risks

Spending risks
5.19 Reflecting the dramatic impact of the pandemic on public spending, of the 27 spending
risks identified in our register (after consolidating three from the original 30 into other risks),
17 have been affected by the coronavirus pandemic in either the medium or long term. Of
these 27 risks, five have crystallised (all remain active), nine have increased, three have
decreased, two have been resolved, and only eight remain unchanged since 2019. None
se to our 2019 report.

Fiscal risks report 218

 Update on other fiscal risks

Table 5.4: Latest assessment of the spending risks identified in our 2019 report
Probability of Impact on Coronavirus Treasury Change
crystallising PSND impact response in risk
Medium term
Less spending subject to DEL controls* Very high Not quantified Yes ◆
Major project cost overrun* Medium Low Yes ◆
Spending announced outside SRs* Very high Medium Yes ◆
State pension triple lock High Medium Yes 🡅
Implementation of welfare reform Very low Low Yes 🡇
Welfare system legal challenges Low Low No 🡇
Precedent from reversing welfare cuts High Low Yes 🡅
Additional health spending Very high Low Yes 🡅
Topping-up health spending settlements Very high Medium Yes 🡅
Health costs from NLW and migration High Low No 🡅
Adult social care Very high Medium Yes 🡅
Higher tax litigation costs Low Low No ⚊
Higher clinical negligence payments Very low Low Yes 🡅
Clinical negligence: legal fees costs Medium Low No ⚊
LAs running down reserves Low Low Yes 🡇
LAs borrowing for commercial property Very low Low Yes 🞫
Devolved administration borrowing* Medium Low Yes ◆
Devolved administration top-ups* Medium Low Yes ◆
Brexit-related exchange rate volatility Very high Low No 🞫
Long term
State pension triple lock Low Medium No ⚊
Additional health spending: demographic Very high Medium Yes 🡅
Additional health spending: other pressures High High Yes 🡅
Adult social care: ageing Very high Low No ⚊
Adult social care: other pressures Very high Medium Yes ⚊
Sellafield clean-up Low Low No ⚊
Clean-up costs for new nuclear plants Low Low No ⚊
Nuclear decommissioning Low Low No ⚊
Note: Refer to the risk register on our website for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium-term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long-term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.
Treasury response refers to the Chancellor's written statement of 14 July 2020. A blank implies it was not discussed.
Coronavirus impact asks whether the pandemic has materially changed our future assessment of the risk.
*Risk remains active despite crystallising.

◆ Crystallised 🡅 Increased ⚊ Unchanged 🡇 Decreased 🞫Removed

5.20 Of the five risks that have crystallised:

Two relate to the control of spending within departmental limits. The risk around the
declining proportion of total spending subject to relatively firm DEL controls has clearly

219 Fiscal risks report

Update on other fiscal risks

escalated during the pandemic, due to both the unprecedented size of the

suspension of multi-year planning that it has prompted. This has also led to the
crystallisation of the tendency for major spending policies to be announced outside
Spending Reviews, with a succession of (often large) spending announcements taking
place during 2020.18 Chapter 2 also considers the significant potential unfunded
legacy costs of the pandemic for public services, focusing on those relating to health,
education and transport. Both issues clearly remain potential fiscal risks for the future.

Two relate to spending by the devolved administrations (DAs). The larger of the two is
the pressure to top-
between the UK Government and the
Scottish and Welsh Governments respectively. This was significantly affected by the
pandemic and the large associated increase in UK Government spending, which had
. In light of the speed and
scale of in-year spending announcements, the Treasury guaranteed additional funding
to the DAs, which by December had reached £16.8 billion that the DAs could spend in
2020-21. By for 2020-21, the Barnett
consequential on additional UK Government spending had reached £18.9 billion, and
the DAs were allowed to choose whether to receive their portion of the additional £2.1
billion in 2020-21 or 2021-22, with all choosing the latter. The guarantees provided a
firmer base from which each DA could plan their spending but, since the amounts
were fixed, they reduced the extent to which it remained directly linked to UK
Government decisions and Treasury spending controls. The second devolved risk that
has crystallised relates to the increased borrowing powers for the devolved
administrations 2021 economy forecast
Scotland-

error from £300 million to £600 million, and will apply from 2021-22 to 2023-24.
Both risks remain ongoing despite crystallising.

One relates to the possibility of cost overruns for major projects, which has crystallised
in several areas that we looked at in our 2019 report. Two large transport projects
High Speed 2 and Crossrail have been affected by the pandemic, with delays and
social distancing requirements lowering productivity and raising construction costs.19 A
third, non-construction, item is the Ministry of Defence Equipment Plan, where the
NAO recently stated that, for the fourth successive year, the Equipment Plan remains
unaffordable .20 Cost overruns for major projects remains a live risk, with the
pandemic requiring new major projects to be set up and their often-large costs to be
managed, notably the very large initial and continuing cost of NHS Test and Trace that
is also discussed in Chapter 2.

18
In 2020-21 there were in effect 14 mini-Budgets in the run-up to the full Budget on 3 March (see Box 3.1 of our March 2021 EFO).
19
Even before the pandemic the NAO reported that there were risks that HS2 costs, already far beyond initial estimates, could rise further.
NAO, High Speed Two: A progress update, January 2020.
20
NAO, The Equipment Plan 2020-2030, January 2021.

Fiscal risks report 220

 Update on other fiscal risks

5.21 Nine risks have worsened since our previous assessment. These are:

Four risks related to higher health spending, both in the medium and long term. These
relate to the consequences of the pandemic for NHS productivity, backlogs of elective
procedures, and the mental health consequences of the pandemic and lockdowns,
among other factors (as we describe in detail in Chapter 2).

The state pension triple lock, in the medium and long term, where earnings growth is
currently particularly uncertain due to base effects (stemming from year-on-year
comparisons being made relative to the initial lockdown-related hit to earnings in
2020), and to compositional effects owing to net job losses being concentrated among
the lower paid, raising the average earnings of those still in work. These factors have
lifted earnings growth to 5.6 per cent in the three months to April 2021, with the Bank
of England noting that earnings growth could rise to 8 per cent over the next two
months due to base effects alone (i.e. even if earnings remained flat at their April level,
due to the weakness in the same months last year).21 The triple lock raises spending by
£0.9 billion a year for every 1 percentage point, and our March forecast assumed
uprating of 4.6 per cent next April. So, if earnings growth in the three months to July
period that determines triple lock uprating for next April were 8 per cent, that would
add around £3 billion a year to spending relative to our forecast. The ratchet effect of
the triple lock means the higher starting point would raise state pensions costs relative
to GDP in all future years too. Over the long term, the downward revision to
productivity growth we made in March 2020 would imply the 2.5 per cent minimum
uprating being triggered more frequently, which would again raise state pensions
spending progressively as a share of GDP each time it happened.

Risks related to the uncertain medium-term costs of adult social care, including around
, and the potential
pressure to bail out a private social care provider in financial difficulty. The policy
uncertainty around medium- and long-
care costs has barely moved in the ten years since the Dilnot Review, but costs for
social care providers have risen during the pandemic, reflecting factors like purchasing
PPE and implementing social distancing. At the same time, the tens of thousands of
excess deaths in care homes over the past year will have reduced incomes.
Overall this has increased the fiscal risk associated with spending on social care.

The precedent set by yielding to pressure to reverse planned cuts to welfare spending.
This risk tends to be more acute where there are clear and identifiable cash losers.22 A
recent example is the extension of the £20-a-week increase to the universal credit
standard allowance in the March Budget, which was initially due to expire in April
2021 and is now slated to end this September. The uplift is now due to be withdrawn
around the same time that the furlough scheme ends, which could be associated with
rises in unemployment, so pressure to extend the policy could build again. The six-

21
Bank of England, Monetary Policy Summary and minutes of the Monetary Policy Committee meeting ending on 22 June 2021, June
2021.
22
As discussed in our December 2019 Welfare trends report.

221 Fiscal risks report

Update on other fiscal risks

month extension that was announced in the March Budget cost £2.2 billion (with an
equivalent payment for working tax credit recipients costing a further £0.8 billion).

The likelihood of higher clinical negligence pay-outs than currently provisioned for,
where the Government extended litigation cover to more providers in the pandemic.

5.22 Three risks have improved in the past two years:

Risks surrounding the implementation of the new state pension and universal credit,
where DWP systems coped well with the surge of 3 million new claims that were started
between 16 March and 31 May last year during the first lockdown.23

The risk around limited formal reporting of the cost of potential legal challenges to the
welfare system Annual Report
& Accounts. The disclosed amount does, however, seem narrowly defined.

The possibility that local authorities will resume running down their reserves has also
decreased, despite the severe pressures on their finances due to the pandemic, as
more unexpected costs having been borne centrally. In terms of risks to the public
finances as a whole, this means the risks associated with such costs crystallise via
central rather than local government. It does not reduce fiscal risk overall.

5.23 Finally, there are two risks that have been resolved:

The ,
where the pandemic has reduced the attractiveness of these investments while the
Government also tightened the rules in November 2020, making it harder for local
authorities to use the Public Works Loan Board in this way. The combined effect of
these developments led us to revise down local authority capital spending from these
sources by £2.6 billion a year on average from 2021-22 onwards.

The exposure to potentially greater exchange rate volatility as a result of Brexit, where
there was little exchange rate volatility around the actual departure date, particularly
relative to the large and sustained fall at the time of the referendum in June 2016.

Balance sheet risks

5.24 Of the eleven balance sheet risks we identified two years ago, four have crystallised (though
three remain active), one has increased, one has decreased, two are unchanged, and three
have been consolidated into others
to our 2019 report.

23
See our March 2021 Welfare trends report for further detail.

Fiscal risks report 222

 Update on other fiscal risks

Table 5.5: Latest assessment of the balance sheet risks identified in our 2019 report
Probability of Impact on Coronavirus Treasury Change
crystallising PSND impact response in risk
Medium term
Public sector net worth* Medium Not quantified Yes ◆
Asset sales* Low Low Yes ◆
Guarantees in infrastructure and housing* Very low Medium No ◆
Housing associations N/A N/A No ◆
Reclassifications and balance sheets Medium Medium Yes 🡅
PSND and fiscal illusions Very high Medium No ⚊
Long term #N/A
Contingent liabilities N/A N/A Yes 🡇
Balance sheet management N/A N/A No ⚊
Note: Refer to the risk register on our website for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium-term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long-term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.
Treasury response refers to the Chancellor's written statement of 14 July 2020. A blank implies it was not discussed.
Coronavirus impact asks whether the pandemic has materially changed our future assessment of the risk.
*Risk remains active despite crystallising.

◆ Crystallised 🡅 Increased ⚊ Unchanged 🡇 Decreased 🞫Removed

5.25 The crystallised risks are:

Public sector net worth has deteriorated significantly due to the costs associated with
the pandemic, which remains an ongoing risk. One aspect of the deterioration over
the past two years, unrelated to the pandemic, is the Governme
- public sector pension ruling where Government estimates suggest
a total balance sheet cost of £17 billion, with the associated spending spread over the
next 60 to 70 years.24 This has yet to be reflected in official estimates or our forecasts.

Asset sales expected to yield £2.6 billion in 2024-25 have been delayed, as we
highlighted in our March EFO, and there remains the risk of further delays or the sales
raising less than expected. Future risks are now focused on t
sell its remaining stake in NatWest Group (formerly RBS), which our March forecast
assumes will raise £13 billion over the next five years. Since then it has completed two
sales, on 19 March and 11 May, raising £1.1 billion from each.

The growing size of guarantees in infrastructure and housing, with the introduction of
the UK Investment Bank (UKIB) in the March Budget and the new 95 per cent
mortgage guarantee scheme that was introduced in April. The UKIB can issue £10
billion of guarantees so this is another risk that remains ongoing.

24
See Box 3.5 of our March 2021 Economic and fiscal outlook. The original £17 billion estimate is from: Public service pension schemes:
changes to the transitional arrangements to the 2015 schemes, Consultation, HM Treasury, July 2020. The Public Accounts Committee

wing in public pension costs

223 Fiscal risks report

Update on other fiscal risks

The risk relating to the statistical classification driving regulatory policy in respect of
housing associations. Following the ONS decision in 2020 to reclassify housing
associations from the public to the private sector, which in turn followed Government
changes to regulations that were expressly designed to relinquish sufficient control over
housing associations to change their statistical classification,25 we have removed the
risk from the register. That said, the broader fiscal risk associated with housing
associations as the vehicle to deliver
remains.

5.26 The risk that has increased is that associated with statistical reclassifications, due to
Government interventions to support different activities during the pandemic, as well as the
creation of the UKIB. These could crystallise into the reclassification of currently private
sector entities to the public sector. Indeed, one already has, with some train operating
companies brought into the public sector during the pandemic. The Government has also
converted loans into equity stakes in several start-up businesses, which could eventually be
considered controlling interests for statistical purposes. The recent announcement of the
creation of the new public body, Great British Railways, could also expand the public

5.27 The risk that has decreased relates to the management of contingent liabilities. In April the
Government launched its Contingent Liability Central Capability in UK Government
Investments to strengthen contingent liability expertise and risk management across
government. Its remit includes analysing and reviewing both new and existing contingent
liabilities the latter being the key information gap that we identified two years ago.

5.28 The risk associated with the use of PSND as a fiscal sustainability metric and fiscal illusions
remain unchanged overall, though this is due to offsetting factors. Fiscal illusions relating to
student loans have now been removed thanks to changes in accounting treatment in the
official statistics that better match economic reality. The ONS also began publishing more
data on public sector net worth capturing a broader range of assets and liabilities. But these
improvements are offset by the risks that might arise from the UK no longer being part of
the European Statistical System and so losing the external oversight and audit function
previously provided by Eurostat. This might encourage greater exploitation of statistical
boundaries.

Fiscal policy risks

5.29 Understanding the way governments make fiscal policy and react to events is important
when assessing the future sustainability of the public finances. Of the three risks we
identified in our 2019 FRR, two have increased and one decreased:

The tendency to revise fiscal rules in line with movements in the forecast. The fiscal
rules in the existing Charter for Budget Responsibility have now expired, with the

25
As discussed in our November 2017 Economic and fiscal outlook.

Fiscal risks report 224

 Update on other fiscal risks

set out in the Con

allow for higher borrowing than the legislated rules. The Chancellor stated his
intention to set out new fiscal rules later this year, conditional on the economic
circumstances. This risk has therefore increased.

The tendency to respond asymmetrically to movements in our underlying forecasts.

This risk has decreased. Presented with a structural deterioration in our medium-term
fiscal forecast due to the pandemic, the Chancellor chose to tighten medium-term
fiscal policy rather than absorbing it in higher borrowing, through reductions in
spending and increases in corporate and personal taxes over the medium term.

Assuming cuts outside Spending Review periods, but then revising totals up when plans
are set. The first element of this risk has been aggravated by the £14½ billion a year
cuts to pre-pandemic departmental spending totals from 2022-23 onwards. Given the
pandemic-
looks particularly challenging, increasing the risk that totals are raised when detailed
departmental spending allocations are set later this year.

Table 5.6: Latest assessment of the fiscal policy risks identified in our 2019 report
Probability of Impact on Coronavirus Treasury Change
crystallising PSND impact response in risk
Medium term
Fiscal rules moved in line with forecast High Low Maybe 🡅
Respond asymmetrically to forecast changes Medium Low Maybe 🡇
Post-SR spending assumptions do not hold High Medium Yes 🡅
Note: Refer to the risk register on our website for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium-term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long-term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.
Treasury response refers to the Chancellor's written statement of 14 July 2020. A blank implies it was not discussed.
Coronavirus impact asks whether the pandemic has materially changed our future assessment of the risk.
*Risk remains active despite crystallising.

◆ Crystallised 🡅 Increased ⚊ Unchanged 🡇 Decreased 🞫Removed

Risks we have removed or replaced for other reasons

5.30 Alongside the updated risk assessments described above, we have consolidated the
presentation of several risks in the register. In the process, we have removed or replaced the
remaining 21 risks that were included in our 2019 risk register (Table 5.7):

Five of those relate to the cost of public debt, where the 2019 list has been replaced
with a new list of risks that reflects the analysis and scenarios presented in Chapter 4.

225 Fiscal risks report

Update on other fiscal risks

The six climate change risks we included in 2019 were conceptual issues around how
to consider climate-related fiscal risks. These have been replaced with three risks
focused on different paths for climate change and for the transition to net zero
emissions in the UK, underpinned by three of the scenarios presented in Chapter 3.

Nine risks relate to questions for the Government about wider risk management. These
were first raised in our 2017 report
response. They are pertinent to the management of many risks across the register and
remain important, but they cannot be quantified in their own right. We have therefore
removed them from the register so that it focuses on the underlying risks themselves.

The final risk we have removed is the stress test that we carried out in the 2017 FRR,
which provides an illustration of several specific risks occurring at once (in this case a

Table 5.7: Risks removed or replaced since 2019

Cost of public debt

Increase in the debt stock and the issuance of index-linked gilts in recent years
Balance between inflation risk exposure and other goals in ILG issuance choices
Potential fiscal impacts of material changes to the Retail Prices Index
Temporary impact of the APF in lowering the Government's borrowing cost
Balance of risks around the future path of the growth-corrected interest rate
Climate change
Integration of climate-related risks into the broader management of fiscal risks
Appropriateness of the Bank/NGFS scenario framework for assessing fiscal risks
Analysis of the sources and transmission channels relevant to the public finances

Way to manage potential shocks to the public finances from climate change
Trade-off between longer-term climate-related fiscal pressures and other priorities
Risk management
The need to review risks governments choose to expose themselves to
The need to prepare for near-inevitable future shocks
The need to deal with many slow-building pressures
Challenges of dealing with those needs while negotiating Brexit
Challenges of doing so in an environment of apparent 'austerity fatigue'
More vulnerable starting position from which all of this is faced
Sources of fiscal risk that we have not analysed - major wars and climate change

How to ensure that the impacts of these changes can be meaningfully assessed?

Other
Fiscal risks report 2017 stress test: severe recession

Fiscal risks report 226

 Update on other fiscal risks

Additions to the risk register as a result of this report

5.31 The analysis in Chapters 2 to 4 has led us to add a further 15 risks to our risk register, nine
relating to the coronavirus pandemic, three from climate change, two from the cost of
public debt, and a final one on the fiscal risks posed by cyberattacks. Some of these are
genuinely new risks that have emerged over the past two years, while others are longer-
standing risks that we have assessed for the first time.

Coronavirus pandemic risks

5.32 As discussed in Chapter 2, the coronavirus pandemic has had immediate and wide-ranging
, creating a legacy of potential medium- and long-term
risks. We have added the following 9 risks to our register:

The risk of future pandemics. The number of infectious disease outbreaks around the
world has risen significantly in recent decades (as discussed in Chapter 1), culminating
in the coronavirus pandemic. The economic and fiscal risks that have crystallised in the
UK as a result of this pandemic are discussed from paragraph 2.3.

The post-pandemic pressures on public services. This generates three new risks, one
each for spending on health and social care, transport and education. These are
discussed from paragraph 2.40.

Risks relating to the fiscal cost of guarantees extended by the Government for the
different loan-support schemes. These are discussed from paragraph 2.32.

Risks relating to scarring of potential output. We added four risks under this heading,
one that is overarching, underpinned by individual risks associated with population,
employment rate and productivity. These are discussed from paragraph 2.51.

227 Fiscal risks report

Update on other fiscal risks

Table 5.8: Coronavirus pandemic risks

Probability of Impact on
Timescale crystallising PSND
Coronavirus
Future pandemics Long term Medium Medium
Post-pandemic pressures on public services
of which:
Health and social care Medium term High Medium
Transport Medium term High Low
Education Medium term High Low
Government guaranteed loans Medium term Medium Low
Scarring of potential output
Pandemic scarring of potential output
of which:
Population Medium term Medium Medium
Employment rate Medium term Medium Medium
Productivity Medium term Medium Medium
Fiscal legacy of the pandemic
Structural shift in receipts accelerated by pandemic Medium term Low Low
Note: Refer to the risk register on our website for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.

Climate change risks

5.33 Chapter 3 considers the fiscal risks posed by climate change, and the economic and fiscal

legislated target to reduce net greenhouse gas emissions to zero by 2050. To construct
paths for these effects, we draw on scenarios produced by the Climate Change Committee
(CCC) for whole economy costs and savings from decarbonisation, and by the Bank of
England for the price of carbon necessary to achieve net zero and its economic implications.

5.34 Based on this analysis, we have added the following three risks to our register:

Unmitigated climate change. If the world fails to bring global warming under control,
physical risks from higher temperatures in the UK, and the consequences of spillovers
from greater impacts in hotter countries, could be very fiscally damaging. Our
illustrative scenario sees debt rising to 289 per cent of GDP by the end of the century.
This high impact risk is considered low probability thanks to the progress being made
under the Paris Agreement, with 131 countries now committed to achieving net zero.

Early and smooth action to achieve net zero, with policy measures to offset predictable
receipts losses. If the measures necessary to achieve net zero are put in place promptly
and smoothly both globally and in the UK the economic and fiscal cost of the

Fiscal risks report 228

 Update on other fiscal risks

transition could be modest. If, in addition, the Government maintains the tax burden
on motoring as fuel duty receipts fall away with the switch to electric vehicles, the
largest fiscal cost of the transition would be ameliorated. Our early action scenario
with motoring taxes maintained actually sees debt reaching 3 per cent of GDP below
the baseline in 2050-51, thanks to the carbon tax revenues that both incentivise
decarbonisation in the private sector and also fund public spending on the transition.

Late and disruptive action to achieve net zero. If action globally and in the UK to
achieve net zero by 2050 were delayed another decade, but then imposed abruptly so
that households, businesses, and financial markets could not adjust smoothly, the
economic and fiscal consequences of the transition would be more costly. Our late
action scenario sees debt reaching levels in 2050-51 that are 47 per cent of GDP
higher than the early action scenario with motoring taxes maintained.

Table 5.9: Climate change risks

Probability of Impact on
Timescale crystallising PSND
Scenario analysis
Unmitigated climate change Long term Low High
Early and smooth transition to net zero Long term Medium Low
Late and disruptive transition to net zero Long term Medium Medium
Note: Refer to the risk register on our website for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.

Cost of public debt risks

5.35 Chapter 4 presents scenarios for the fiscal consequences of possible future paths for interest
rates. We have used these, plus the latest balance sheet data, to refresh our presentation of
debt interest risks and to look at the consequences of tail risks around investor confidence.
Three risks have been retained from the 2019 register and two have been added:

Higher stock of public debt increases sensitivity to rate changes. Chart 4.8 shows that
the higher stock of debt means that a 1 percentage point rise in interest rates will
ultimately increase debt interest spending by 1.1 per cent of GDP, three times as much
as it would have in 2007-08, just prior to the financial crisis.

Reduced median maturity of consolidated public sector liabilities increases the speed
of pass-through of interest rate rises. Chart 4.8 also shows that the increase in debt
interest spending within the first year of an increase in interest rates has risen six-fold
over the same period, including a sharp increase in 2020, reflecting a by-product of
quantitative easing that raises the proportion pass-through that happens immediately.

229 Fiscal risks report

Update on other fiscal risks

The impact of higher inflation on government debt. The fiscal impacts of higher
inflation are explored from paragraph 4.56. Our scenarios show that a temporary
shock may benefit debt as a share of GDP, but a permanent rise would not.

The impact of higher real interest rates on government debt are explored from
paragraph 4.39. Our scenarios show how the impact on debt depends on the extent to
which rate rises are accompanied by higher economic growth. If they are, the growth-
corrected interest rate is little changed and fiscal implications are modest; if they are
not, and the growth-corrected interest rate rises, higher debt interest spending is not
offset by higher revenue growth, with adverse fiscal implications.

A loss of investor confidence in UK sovereign debt. This tail risk scenario results in a
risk premium on government borrowing costs and a recession that together generate a
debt spiral and loss of fiscal sustainability. It is discussed from paragraph 4.68.

Table 5.10: Cost of public debt risks

Probability of Impact on Change
crystallising PSND in risk

Size and composition of debt (Medium term)

Higher stock of debt increases sensitivity to rate changes Crystallised Medium Added
Reduced median maturity increases sensitivity to rate rises Crystallised Medium Added

Sensitivity to inflation and interest rate risk (Long term)

e impact of higher real interest rates on government debt Medium Medium 🡅
e impact of higher inflation on government debt Medium Medium ⚊
loss of investor confidence in UK sovereign debt Very low High 🡅
Note: None were referenced in the Chancellor's written statement of 14 July 2020. Refer to the online risk register for more details.
Medium term is within 5 years and long term beyond that.
Probability: Very Low = <10%; Low = 10%-40%; Medium = 40%-60%; High = 60%-90%; Very High = > 90%.
Medium-term impact (using 2025-26 GDP): Low = <1% of GDP; Medium = 1%-10% of GDP; High = 10%-100% of GDP.
Long-term impact (using 2070-71 GDP): Low = 1%-10% of GDP; Medium = 10%-100% of GDP; High = >100% of GDP.

◆ Crystallised 🡅 Increased ⚊ Unchanged 🡇 Decreased 🞫Removed

Cyberattacks
5.36 The final addition to our 2021 fiscal risks register is the risk posed by a cyberattack with
systemic consequences that causes sufficient disruption to have macroeconomic and fiscal
implications. At this stage we have not quantified the potential impact of such an event,
which we will attempt to do in our next FRR.

Fiscal risks report 230

 Update on other fiscal risks

Box 5.1: The fiscal risks posed by cyberattacks

cyberattack. Cyberattacks are a growing threat, but to date none have caused sufficient
disruption to critical national infrastructures to have caused material economic and fiscal harm.a

But the relatively small scale of the damage of cyberattacks to date may not be a good guide to
the risk of more significant harm being done in the future. They have been on a sharply rising
trend (left panel of Chart A). On one measure, the UK is ranked in the top ten countries in the
world in terms of global connectedness, so is arguably more vulnerable to cyberattacks by virtue
of its role as a major global financial centre and the international reach of many of its
companies.b
of significant cyberattacks between 2006 and 2020, behind only the US (right panel).c
Chart A: Significant cyberattacks since 2006
160 160
Globally by year Cumulatively by country
Number of significant (losses >$1m)
Number of significant (losses >$1m)

140 140

120 120
cyber incidents
cyber incidents

100 100

80 80

60 60

40 40

20 20

0 0
2006 2008 2010 2012 2014 2016 2018 2020 US UK Ind Ger Kor Aus Ukr Chi Iran Sau
Source: Centre for strategic and international studies Source: Specops

National Risk Register places

millions rather than billions of pounds. It warns that cyberattacks can impact critical national
services, and could cause a variety of real-world harm if services like the NHS are impacted . The
latter crystallised albeit modestly
seven days of disruption across one-third of hospital trusts at a cost of £92 million.d
Cyberattacks come from a variety of sources including criminal and terrorist organisations,
-sponsored activities. The Chief Executive of the National
Cyber Security Centre (NCSC) has warned that state actors have been a constant presence in
recent years, but that for the vast majority of UK citizens and businesses, and indeed for the vast
majority of critical national infrastructure providers and government service providers, the primary
threat is not state actors but cyber criminals, and in particular the threat of ransomware .e
The number of ransomware attacks has increased in recent years. On one estimate, over $400
million of payments were made by ransomware victims in 2020, with growth in recent years
having been exponential.f In the UK, the NCSC reports that it handled three times more
ransomware incidents in 2019-20 than in the previous year.g

231 Fiscal risks report

Update on other fiscal risks

Some recent attacks illustrate the potential for wider economic and fiscal consequences, though
they were resolved before such effects crystallised. These include disruption to fuel supplies
across parts of the US that could have resulted from the attack on the largest fuel pipeline in the
US by the group DarkSide, and the hack on the US company SolarWinds, where malicious code

the world, the consequences of which may not be fully understood for many years.
Future cyberattacks could pose a major threat to the functioning of the global financial system,
with an attack on one institution potentially spreading rapidly to others. To that end, the Bank of
England is undertaking a cyber stress test for UK financial institutions in 2022.h Such attacks
could pose material macroeconomic and fiscal risks. An IMF study estimates that average annual
losses from cyberattacks on the financial system could be in the region of $100 billion globally,
and in more severe scenarios might reach as high as $350 billion.i

The pandemic has also emphasised our reliance on digital technologies, which facilitated the
rapid switch to working from home for large parts of the workforce, the accelerated shift to
d
degrees of fiscal support to households and businesses, and rapid processing of welfare claims.
So while cyberattacks to date have had modest economic and fiscal implications, it is clear that
they could pose a more material risk in the future. These could manifest themselves via some
combination of: (i) disrupting public services; (ii) disrupting the collection of revenue or payment
of benefits; (iii) disrupting payment systems or threatening financial stability, forcing government
to step in and insure against or meet associated costs; and/or (iv) disrupting the critical national
infrastructure on which the economy depends, like the power grid and transport network. These
could result in various direct and indirect fiscal costs pushing debt higher.
As with our assessment of the fiscal risks from climate change in this report, it may be possible to

more fully in our next Fiscal risks report.

a
Four in ten businesses, including two-thirds of large businesses, reported a cyber security breach in the 12 months to 24 March,
with an estimated average cost of £8,460 each (£13,400 for large businesses). Among these, 27 per cent reported cyberattacks at
least once a week. See, Department for Digital, Culture, Media & Sport, Cyber Security Breaches Survey 2021, March 2021.
b
Global Connectivity Index, Country rankings, January 2021.
c
World Economic Forum, The Global Risks Report 2021, January 2021, originally in Specops, The countries experiencing the most
-attacks, July 2020, and based on Center for Strategic and International Studies, Significant cyber incidents since
2006, June 2021.
d
Department of Health and Social Care, Securing cyber resilience in health and care, October 2018.
e
Lindy Cameron, Speech to the Royal United Services Institute, June 2021.
f
Chainalysis, Ransomware 2021: Critical Mid-year Update, May 2021.
g
National Cyber Security Centre, Annual Review, August 2020.
h
Bank of England, Financial Policy Summary and Record of the Financial Policy Committee Meeting on 11 March 2021, March 2021.
This supplements the ce of individual firms.
i
Christine Lagarde, Estimating cyber risk for the financial sector, June 2018.

Fiscal risks report 232

Index of charts and tables

Executive summary
Chart 1: Incidence of major risk events............................................................................ 4
Chart 2: Fall in real GDP in 2020 in advanced economies ............................................... 5
Chart 3: Fiscal rescue packages across major economies ................................................. 7
Chart 4: Illustrative estimates for selected pandemic-related pressures on departmental
resource spending .............................................................................................. 9
Chart 5: Early action scenario: impact on public sector net debt ..................................... 12
Chart 6: Climate scenarios: impact on public sector net debt in 2050-51........................ 14
Chart 7: Cost of public debt scenarios: public sector net debt ......................................... 16
Chart 8: Sensitivity of interest payments to a rise in interest rates..................................... 17
Chart 9: OBR fiscal risk register: changes since our 2019 report ..................................... 19
Figure 1: Sources of fiscal risk over the medium term ..................................................... 20
Figure 2: Sources of risk to fiscal sustainability ............................................................... 20

Chapter 1 Introduction
Chart 1.1: Battle-related deaths in state-based conflicts.................................................. 27
Chart 1.2: Numbers of international disasters and infectious disease outbreaks............... 28
Chart 1.3: Global connectedness .................................................................................. 29
Chart A: Estimates of government debt limits, thresholds, and targets ............................. 30
Chart 1.4: Public sector net debt ................................................................................... 32

Chapter 2 Coronavirus pandemic

Chart 2.1: World GDP growth and the proportion of economies with falling output ......... 39
Chart 2.2: Annual GDP growth since 1701 ................................................................... 40
Chart 2.3: Sectoral peak-to-trough falls in 2008 and 2020 ............................................ 41
Chart 2.4: The UK economy has adapted over successive lockdowns .............................. 42
Chart 2.5: UK GDP recoveries: pandemic outturn versus forecast and financial crisis ....... 43
Chart 2.6: Real GDP loss versus contributing factors: cross-country comparisons ............. 44
Chart 2.7: Public sector net borrowing and net debt since 1900 ..................................... 45
Chart 2.8: Pandemic borrowing was driven more by spending than receipts ................... 46
Chart 2.9: Increases in government borrowing in selected advanced economies in 2020 . 47

233 Fiscal risks report

 Chart 2.10: International comparisons of health system capacity .................................... 48
Chart 2.11: Fiscal support by recipient for selected advanced economies ........................ 49
Chart 2.12: Fiscal support by instrument for selected advanced economies ..................... 50
Chart 2.13: Percentage change in different tax streams from 2019-20 to 2020-21 ......... 51
Chart 2.14: Contributions to the strength of receipts in 2020-21 .................................... 53
............................................. 54
Table 2.1: Change in spending between 2019-20 and 2024-25 .................................... 55
Chart 2.16: Loans issued under the pandemic-related loan guarantee schemes .............. 57
Chart 2.17: Loans approved and expected fiscal costs of loan guarantees ...................... 58
Chart 2.18: Illustrative estimates for selected pandemic-related pressures on departmental
resource spending ............................................................................................ 60
Chart 2.19: NHS waiting lists ........................................................................................ 63
Chart 2.20: Changes in public transport mobility during the pandemic ........................... 65
Chart 2.21: Business investment during the pandemic .................................................... 67
Chart 2.22: Contributions to change in participation between 2020Q1 and 2021Q1 ...... 69
Chart 2.23: Comparison of GDP forecasts .................................................................... 71
Table 2.2: Forthcoming scarring-related data releases ................................................... 73
Chart A: Change in employment-related indicators during the pandemic ........................ 74
Chart B: Sectoral breakdown of furloughed employees and vacancies (May 2021) .......... 75
Chart 2.24: Coronavirus deaths versus net fiscal costs by age group .............................. 76
Table 2.3: Potential long-term impacts on the receipts-to-GDP ratio................................ 78

Chapter 3 Climate change

Chart 3.1: Carbon dioxide emissions and atmospheric concentration ............................. 86
Chart 3.2: Median temperatures versus the risk of exceeding 6°C temperature rises at
different atmospheric concentrations of greenhouse gases .................................. 87
Chart 3.3: Estimated impact on emissions of the existence of different interventions ......... 94
Chart 3.4: Global greenhouse gas emissions in 2016: top 20 emitters ........................... 95
Chart 3.5: The effect of global emissions paths on UK temperatures ............................... 98
Chart 3.6: Public sector net debt: an illustrative unmitigated global warming scenario ..... 99
Chart 3.7: UK greenhouse gas emissions: territorial versus consumption basis .............. 100
Chart 3.8: Cross-country territorial CO2 emissions since 1990 ..................................... 101
Chart 3.9: Reduction in territorial CO2 emissions between 1990 and 2019 .................. 102
Chart 3.10: OECD environmental policy stringency index for the UK ............................ 103
Table 3.1: Selected assumptions on ................... 105

Fiscal risks report 234

 Chart 3.11: CCC balanced pathway for reduction in emissions by sector ...................... 106
...... 107
Chart 3.13: Whole of economy transition costs by sector over 30 years ......................... 108
Chart 3.14: Surface transport: emissions reductions and whole economy net costs ........ 110
Chart A: New car registrations and total car stock by power type .................................. 111
Chart B: Successive assumptions for electric vehicles as a share of new car sales ........... 112
Chart 3.15: Buildings sector: emissions reductions and whole economy net costs .......... 114
Chart 3.16: Electricity supply: emissions reductions and whole economy net costs.......... 119
Chart 3.17: Industry sector: reduction in emissions and whole economy net costs .......... 121
Chart 3.18: Negative emissions and whole economy costs from the removals sector ..... 122
Table 3.2: The share of costs borne by public spending................................................ 126
Chart 3.19: Costs to public sector of the transition to net zero ...................................... 127
Chart 3.20: Loss of motoring, aviation and waste revenues in the balanced pathway..... 129
Chart 3.21: Motoring, aviation and waste tax revenues under alternative scenarios ....... 130
Chart 3.22: Real-terms carbon tax rates: outturn and scenario assumption ................... 131
Chart 3.23: Additional carbon tax revenues................................................................. 132
Chart 3.24: Total direct impact of the transition to net zero on receipts ......................... 133
Chart 3.25: Scenario assumptions: global carbon price, emissions and temperature ..... 135
Chart 3.26: Early action scenario: real GDP ................................................................ 136
Chart 3.27: Early action scenario: key fiscal aggregates ............................................... 138
Chart 3.28: Early action scenario: differences from the baseline debt-to-GDP ratio........ 139
Chart 3.29: Early action versus late action scenarios: fiscal aggregates ......................... 144
Chart 3.30: Alternative productivity variants: real GDP and the debt-to-GDP ratio ......... 145
Chart 3.31: Net zero public spending variants: spending and debt-to-GDP ratios.......... 146
Chart 3.32: Public investment additionality sensitivities ................................................. 147
Chart 3.33: Motoring tax sensitivities ........................................................................... 148
Chart 3.34: Long-term policy assumption sensitivities: debt-to-GDP ratio ...................... 148
Chart 3.35: Contributions to PSND scenario differences from baseline.......................... 149
Table 3.3: Summary of climate-related scenario assumptions ....................................... 150

Chapter 4 Cost of public debt

Chart 4.1: Debt to GDP, the growth-corrected interest rate and net interest payments .... 154
Chart A: Face versus market value of gilts held by the private sector ............................. 158
Chart 4.2: 10-year nominal government bond yields and nominal GDP growth ............ 160
Chart 4.3: Selected high-quality government bond yields and the return on capital........ 161

235 Fiscal risks report

 Figure A: Determination of rates of return.................................................................... 163
Chart 4.4: Middle versus old-aged population globally and in China............................ 165
Chart 4.5: Income inequality and saving rates across the income distribution ................ 166
Chart 4.6: Holdings of selected high-quality government debt ...................................... 169
Chart B: The impact of past pandemics and wars on interest rates ................................ 170
Chart 4.7: Decomposition of the fall in global real interest rates ................................... 171
Chart 4.8: Sensitivity of interest payments to a rise in interest rates ................................ 174
Chart 4.9: Higher global real interest rate scenarios: key assumptions .......................... 177
Chart 4.10: Higher global real interest rate scenarios: key outputs................................ 179
Chart 4.11: Higher global real interest rate scenarios: contributions to differences in debt-
to-GDP ratios from the baseline....................................................................... 179
Chart 4.12: Higher inflation scenarios: key outputs ...................................................... 183
Chart 4.13: Temporary inflation shock impact on net interest payments and PSND ........ 183
Chart 4.14: Persistent rise in inflation impact on net interest payments and PSND .......... 185
Chart 4.15: Long-term bond rates in selected government debt crises ........................... 187
Chart C: Distribution of changes in the UK debt-to-GDP ratio ....................................... 188
Chart 4.16: General government net debt and gross financing needs, 2021 ................. 190
Chart 4.17: General government net worth for selected advanced economies ............... 191
Chart 4.18: UK year-on-year change in the primary surplus ......................................... 193
Chart 4.19: UK year-on-year change in the primary surplus and nominal GDP growth
excluding crisis periods ................................................................................... 194
Chart 4.20: G7 year-on-year change in primary deficit ................................................ 194
Chart 4.21: UK gilt holders ......................................................................................... 196
Chart 4.22: Foreign holdings of general government debt securities ............................. 197
Chart 4.23: Defined benefit pension schemes asset allocations .................................... 198
Chart 4.24: Interest rates in the loss of investor confidence scenario ............................. 200
Chart 4.25: Loss of investor confidence scenario: key outputs ....................................... 201
Chart 4.26: Loss of investor confidence scenario: differences from baseline .................. 202
Chart 4.27: Primary surplus adjustment required to stabilise the debt-to-GDP ratio........ 203
Chart D: Net savings to the public sector of the APF in our scenarios ............................ 204

Chapter 5 Update on other fiscal risks

Chart 5.1: OBR fiscal risk register: changes since our 2019 report ................................ 208
Table 5.1: Latest assessment of the economy risks identified in our 2019 report ............ 211
Table 5.2: Latest assessment of the financial sector risks identified in our 2019 report ... 213

Fiscal risks report 236

Table 5.3: Latest assessment of the revenue risks identified in our 2019 report .............. 214
Table 5.4: Latest assessment of the spending risks identified in our 2019 report ............ 219
Table 5.5: Latest assessment of the balance sheet risks identified in our 2019 report ..... 223
Table 5.6: Latest assessment of the fiscal policy risks identified in our 2019 report ......... 225
Table 5.7: Risks removed or replaced since 2019 ........................................................ 226
Table 5.8: Coronavirus pandemic risks ....................................................................... 228
Table 5.9: Climate change risks .................................................................................. 229
Table 5.10: Cost of public debt risks ........................................................................... 230
Chart A: Significant cyberattacks since 2006................................................................ 231

237 Fiscal risks report

 Fiscal risks report
July 2021

CCS0521649260
ISBN 978-1-5286-2649-1