Europe's turning point - Technical appendix October 2021 - Deloitte
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Europe’s turning point Technical appendix October 2021
Contents
Europe’s turning point: Technical appendix 2
1. Definitions 4
2. D.CLIMATE modelling 6
3. Policy scenario 20
4. Discounting the future 25
5. Limitations 26
Endnotes 28
Authors 34
Deloitte Economics Institute 35
Acknowledgments 35Europe’s turning point
Europe’s turning point:
Technical appendix
In 1990, the first Intergovernmental Panel on Climate The Deloitte Economics Institute strongly recognises
Change (IPCC) report concluded that human-caused the limitations of integrated assessment modelling
climate change would become apparent, but it could (IAMs) – and determining relationships between GHG
not confirm that it was currently happening. emissions, global surface temperature and economic
impacts. But equally, we recognise that economics
Published in August of this year, the latest IPCC sixth
can provide useful insights for rapid decision-making
assessment report provides the most up-to-date
today, and not discount the pragmatic in pursuit of
physical understanding of the climate system and
the perfect.
climate change. In this assessment, the evidence
is clear that the climate has changed since the In this context, the economic modelling conducted
preindustrial era and that human activities are the in this analysis for research purposes has several
principal cause. objectives and seeks to overcome limitations:
With more data and improved models, the y The results indicate an order of magnitude
assessment gives improved estimates and narrower impact on gross domestic product (GDP) and
ranges compared to the previous assessment. other economic variables over the next 50
Global surface temperature will continue to increase years. These results should not be interpreted
until at least the mid-century under all emissions as predictions or ‘most likely’ estimates of
scenarios considered in the assessment. Global climate change impacts. The modelling instead
warming of 1.5°C and 2°C will be exceeded during provides a consistent framework through which
the 21st century unless deep reductions in carbon to understand the economic difference between
dioxide (CO2) and other greenhouse gas (GHG) possible future worlds – one with climate change
emissions occur in the coming decades. There is impacts and one without. Establishing a long-
greater certainty that with every additional increment term view of the impact – albeit narrowed down
of global warming, changes in extremes will become to precise scenario specifications – enables us
larger – for example, every additional 0.5°C of global to draw conclusions as to trade-offs and the
warming causes distinct increases in the intensity direction of change in economies. This is true for
and frequency of hot extremes, including heatwaves, both high emissions, high temperature increase
heavy precipitation as well as agricultural and pathways and low emissions, lower temperature
ecological droughts in some regions. increase pathways.
Modelling the economic impacts of the physical risks y In the absence of transformation, a pathway
from climate change and the economic impacts of of higher emissions and higher global surface
mitigation and adaptation pathways can be fraught, temperature is considered as our baseline
but not insurmountable. The economics discipline outlook for the world. In this outlook, there will
has spent several decades debating the benefits and be economic damages from climate change.
limitations of the established techniques to derive While there are several probable scenarios
economic estimates. To this day, while there remain as to emissions profiles and corresponding
many uncertainties and technical limitations on what temperature increases, all paths will result in
macroeconomic models can reasonably conclude, some degree of climate change damage. In our
economic techniques have improved to integrate work, we adopt a single, higher-emission and
views of the physical climate and economies and higher temperature future pathway (see section
provide important insights into the choices that can 2.1) that offers the basis for an integrated view
be made to drive prosperity. of chronic physical damages becoming a trend.
2Technical appendix
This baseline outlook seeks to demonstrate costs associated with a choice to not act. To this end,
that choosing a path of no change from current economic analysis of climate change is important
global emissions trends is not costless to the to reframe the debate and inform decision-making
economy.1 Economic growth does not, and today, in full understanding of the limitations of both
will not, occur uninterrupted as and when the science and economics.
climate changes.
Governments, business and communities all need
y In reference to this baseline outlook, a view to accelerate decision-making to decarbonise, and
to the costs and benefits of mitigation and this requires economic analysis that accounts for the
adaptation can more reasonably be considered. climate. If we can’t reframe the starting point – that
To inform this, we model a single pathway inaction comes with significant economic costs –
of economic transformation that decouples then any action on climate change will always appear
emissions intensity from the system of economic as an unreasonable cost to society and economies.
production. This is a view to decarbonisation of
Any economic change will have a cost attached to it
economies that aligns to a near net-zero GHG
– whether that is a change in the climate, or a change
emission profile – compared to preindustrial
to decarbonise. It is about how we understand the
levels – and limits global surface temperature
potential magnitude of those costs, the options to
warming to as close to 1.5°C – well below 2°C
minimise them and how the choices we all make
– by 2050. This decarbonisation pathway has
today determine the extent of them. There is a
many variables attached to it. In our work, we
narrow, and closing, window of time to create a
do not prescribe probability or likelihood to
new engine for sustainable economic prosperity
this pathway, much like the baseline. Rather
while preventing the worst consequences of a
the focus is on the sequencing, pace and scale
warming world.
of economic actions and transformations that
could support economies to decarbonise within There is uncertainty in the economic impacts of
a carbon budget that limits global surface climate change and decarbonisation. But there is
temperature warming to well below 2°C by 2050. high confidence we will regret looking up in 2050 to
face a planet with warming and economic loss all
The goal is in understanding the economic rationale
because we did not try to understand the economic
for acting to avoid increases in global surface
rationale to change.
temperature and unmitigated climate change and
IPCC sixth assessment and relationship to analysis
The IPCC has released the Working Group I contribution to the Sixth assessment report (AR6)
as the most up-to-date physical understanding of the climate system and climate change.
AR6 outlines improved knowledge of climate processes, paleoclimate evidence and the
response of the climate system to increasing radiative forcing (driven by higher greenhouse
gas concentrations). AR6 provides the best estimate of equilibrium climate sensitivity of
3°C, with a narrower range compared to the previous AR5. According to AR6, the global
surface temperature will continue to increase until at least mid-century under all emissions
scenarios considered. Global warming of 1.5°C and 2°C will be exceeded during the 21st
century unless deep reductions in CO2 and other greenhouse gas emissions occur in the
coming decades.
The analysis using the D.CLIMATE model does not take a probabilistic approach to the
baseline and net-zero scenarios. Rather, it models narrow economic impacts that relate
to specific damage functions from the selected emission pathways, without assigning
probability to the outcome. This approach is to inform a better framework for decision-
making today, based on orders of magnitude of economic trade-offs over time.
3Europe’s turning point
1 Definitions
1.1 Net zero and technologies offset the transitions from
emissions-intensive production processes.
The Deloitte Economics Institute has modelled a
scenario that reflects the world reaching net-zero 1.5 Deviations from baseline
GHG emissions by 2050. Of this, around 13.4Gt
Europe’s turning point The turning point narrative is based on scenario
(or around 20 per cent) of CO2e is expected to
analysis. The modelling does not provide a
be offset or captured via carbon sinks in 2050.2
forecast of the future, but rather comparisons
The EU offset or captured emissions are between possible future worlds. The discussion of
benchmarked to feasible LULUCF, natural modelling results will usually describe the state of
and man-made carbon sinks by 2050, noting the economy in reference to an alternative future
that the likelihood of these becoming viable is – the deviation in a variable (like GDP) from one
enhanced by the modelled carbon price. scenario (i.e. transition) to another (i.e. inaction).
1.2 Close to 1.5°C world 1.5.1 Climate change impacts
This pathway to net zero for the world limits When comparing two alternative futures, a lower
global average warming to well below 2°C and GDP at the same point in time is not the same as
close to 1.5°C compared to preindustrial levels, having negative GDP growth. The example below
in alignment with current Paris Agreement demonstrates that comparing two GDP scenarios
objectives. Thethe
In this report, precise warming
Deloitte in global
Economics average
Institute with and without the impacts of climate change
mean
presentssurface temperature
two scenarios. Theused
first in modelling
describes shows that both scenarios reflect an economy
is 1.7°C
what above
could preindustrial
happen levels byof2100.
if the countries the world
Climate
that change
is still growing. Thewill impact
impact or ‘loss’our
of GDP
allowclimatic
the planet economic
in 2022 due tofuture, but a
climate change path
is the to net
difference
The and to warm onimplications
economic a path to 3ºC
of higher
this
temperatures by the end of the century.
global temperature pathway are modelled Theas the zero can
between thecreate
two GDP new opportunities
levels in figure 1 below.
second reveals
comparison the economic
scenario to a worldopportunities for
of climate inaction. for Europe
Europe if the world limits global warming to as
1.3
closeAround
to 1.5°C3°C world
by mid-century. FIGURE 1
An
1. economic scenariofurther
We do nothing that relates
and to a pathway
global Accounting for climate
The impact of accounting forchange impacts
climate change
of climate inaction.
emissions riseThe socioeconomic
(‘around and This
3ºC world’): on Europe’s growth path
emissions pathways
economic underpinning
path represents this exercise
a future with a
are broadly consistent with theemissions,
SSP2-6.0 scenario 2021 2050 2070
higher rate of global GHG where
(see there
section 2.4.1).
are
3
The implied
no significant temperature
additional mitigation Assumed growth
change is 3°C
efforts andabove preindustrial
the global averagelevels by 2100.
temperature GDP growth without
increasespointby around 3°C by 2100. This scenario accounting for
1.4 Turning concept climate damage
reflects a widely adopted set of emissions,
GDP growth path
Transitioning
economictoand a net-zero world
population and limiting
assumptions,
warming to astoclose
referred to 1.5°C requires
as SSP2-6.0. an industrial
This scenario is
and economic
regionalised transformation
to the European thatcontinent
would typically
and
Corrected growth
occurtheover a centurywithin
economies to take
it.place in just of
The results three
this GDP growth once
decades. The are
scenario turning point concept
presented highlights
as a deviation, a climate damage is
that choosing
comparison to rapidly accelerate
to a world that doeswillnot
mean
have that, accounted for
despite initialchange
climate costs, impacts
countriesmodelled.
and industries will
see dividends to this investment. It is a climatic
We act decisively
2. economic
and andinquickly
turning point the sense to hit
thatglobal Source: Deloitte
Source: DeloitteEconomics
EconomicsInstitute.
Institute.
net zero by mid-century (‘close
the worst effects of climate change are avoided to 1.5ºC
world’):
and the economic benefits of new industries a
This economic path represents
sequencing of efforts – by governments,
If we do nothing further, global
4 businesses and citizens – to achieve net-zero
emissions will rise
emissions by 2050. This scenario would make
This illustration depicts what Europe’s grow path
it possible for us to limit warming to as closeTechnical appendix
1.5.2 Deviations from a damaged baseline 1.6 Decarbonisation
Unlike most modelling exercises, though, the Decoupling emissions from growth in our
Economics for acontext.
new climate
deviations presented in the net-zero scenario Emissions intensities do not change by industry,
involve a two-step calculation to account for the but the industrial composition and production
combined impact of avoided damages alongside processes adapt to rely on less emissions-intensive
transition costs. This is done to reflect the idea activities.
that
2021a more appropriate
2050 baseline
2070for modelling
1.7 Clean energy and electricity
and decision-making is one that accounts for
Assumed growth
climate impacts that arise from emissions-intensive Includes solar, wind, nuclear, hydropower and
GDP growth without
economic growth. accounting for
geothermal production technologies. Zero-
climate damage emission hydrogen and bioenergy are included in
In a simple example, a region might be expected
clean energy (see section 2.3.3).
GDP growth path
to lose 3 per cent of output (figure 2) due to the
damages associated with less ambitious domestic/ 1.8 Conventional energy and electricity
global action on climate change. In a scenario
Corrected growth Includes coal, oil and gas as fuels and energy
where they and the rest of thecould
world takeGDP more Rapid, coordinated
If we act decisively, Europe growth once yproduction as well as theirglobal
use indecarbonisation
electricity
hit global net zero by mid-century climateloss
ambitious action, there might be a smaller due
damage is would not only limit the
production. Carbon capture, use worst effects is not
and storage
to damages, but there will be some cost associated
accounted
The net-zero scenario we present in this report for
of climate change, but could bring an
separately modelled.
with the
shows howpolicies
Europeenacted to reduce
could limit climateemissions.
damage The economic and climate turning point:
results presented here show the combination
and achieve new growth. This illustration of Transitioning to a net-zero world and limiting
damages and transition costs, relative
shows that rapid decarbonisation could offer to the more warming as close to 1.5°C requires an industrial
severeeconomic
more damagesbenefits
from thetobaseline scenario.
Europe than our and economic transformation. The turning point
current trajectory, the path of inaction. concept highlights that choosing transition
FIGURE 2 will mean that, despite initial costs, countries
Net-zero scenario impacts compared to a and industries could see dividends in terms of
climate-damaged baseline
The opportunity of new economic growth avoided costs from climate damage and in the
under a net-zero scenario form of new industries and technologies.
2021 2050 2070 y The net gain or the turning point is
Assumed growth different for every economy: The turning
GDP growth without point is different for each economy because
accounting for it reflects the combined impacts of economic
climate damage
costs and benefits, the economic structure that
creates growth today, and how that’s modelled
GDP growth path
Net-zero scenario
Economic impact of is impacted by decarbonisation. It also greatly
decarbonisation depends on the exposure to ‘locked-in’ climate
Corrected growth
change impacts as the world warms by at least
GDP growth once 1.5°C over the coming decades.
climate damage is
accounted for – Costs:
The inevitable costs to the economy
as it moves away from emissions-
Source:
Source: Deloitte
DeloitteEconomics
EconomicsInstitute.
Institute. intensive activity
The cost to the economy from global
warming of at least 1.5ºC, even with strong
global action to reach net zero by 2050
Pull quote?
– Benefits:
The benefit of avoiding costs from limiting
global warming, instead of reaching
around a 3ºC increase in global average
temperatures
The benefit of a more productive and
modern economy, where demand is
being met as consumer and industry
preferences change
5Europe’s turning point
2 D.CLIMATE modelling
2.1 Overview it the functionality of a fully fledged integrated
assessment model (IAM). Unlike many IAMs, this
To date, most macroeconomic models and
model has multiple economic damages that vary
economic policy analysis are considered against a
by sector and region, and unlike many regional
‘baseline’ that assumes economic growth will occur
computable general equilibrium (CGE) models,
unhindered by rising concentrations of GHGs in
it has full integration with the global economy
the world’s atmosphere. The Deloitte Economics
through the Global Trade Analysis Project (GTAP)
Institute believes that this viewpoint does not hold
database and a complete set of emissions
true in practice – particularly in the long run – and
accounts covering CO2 and non-CO2 gases.5
therefore economic analysis and climate policy are
informed through a dated theoretical framework. This work draws on, and contributes
to, three key streams of research:
Climate change impacts should not be considered
as a scenario ‘relative to’ a baseline of unconstrained y The primary stream is that which has
emissions-intensive growth – because in the pioneered, refined and expanded CGE
absence of fundamental societal and economic models, allowing for modelling of complex
shifts, the impacts of unmitigated climate change and dynamic policies, like those required to
are the baseline. By excluding the economic impacts affect a transition to a low-carbon environment
of climate change from economic baselines, (see Adams and Parmenter, 2013).6
decision-making misses a fundamental point.
y Another stream is that which has followed the
A shift to understand and incorporate this same process of pioneering, refinement and
climate-affected baseline into decision-making expansion, but for IAMs.7 The IAM stream,
is gaining momentum. The Network for Central in its initial phases, used a more aggregate
Banks and Supervisors for Greening the representation of the economy that allowed
Financial System, made up of 92 central banks, for a stylised climate module.8 These models
has been a prominent example, developing sought to establish a link between the
guidance and scenarios to assist the financial economic system potential damages associated
sector to better understand its climate risks.4 with climate change to be incorporated
to form an integrated (but simplified)
Understanding and accounting for the longer-
framework for assessing the decisions facing
term effects of climate change on productivity,
policymakers when it came to emissions
potential output and economic growth are
reduction targets (see Nordhaus, 2013).9
critical to understanding the likely future growth
path of the global economy, as well as the y The third and most recent stream is that which
distribution of disruptive climate impacts. seeks to combine the two described above
and provide the rich sectoral and policy detail
The Deloitte Economics Institute has invested in
inherent in modern CGE models, alongside
developing an extension of the in-house regional
climate feedback mechanisms that allow for
general equilibrium model (DAE-RGEM), giving
integrated assessment (see Kompas, 2018).10
6Technical appendix
D.CLIMATE is an extension of a well-established an evolution in atmospheric GHG concentration
modelling methodology and policy analysis that rises in line with a Representative
technique that seeks to ‘correct’ the typical business- Concentrative Pathway (RCP).
as-usual baseline assumed in most modelling.11
y Rising atmospheric concentrations of GHGs
D.CLIMATE is built on an economic modelling causes global warming above preindustrial
framework that accounts for the economic impacts levels, as projected by a reduced complexity
of climate change and establishes a reference climate model (the Model for the Assessment
case that can be modelled out to the year 2100 or of Greenhouse Gas Induced Climate Change,
beyond. The D.CLIMATE process and logic are as MAGGICC).12
follows:
y Warming causes shifts in global climate patterns
y The modelling produces a baseline economic and results in damages to the factors of
growth path that draws on short- to medium- production (capital, labour and land) and their
term global and regional forecasts in productivities.
combination with a long-run assumption of
y Damages to factors of production are
contraction and convergence.
distributed across the economy, impacting GDP.
y The baseline economic growth path has an
y These feedbacks are fed back into the model to
associated emissions growth path – derived
determine the associated deviation in economic
from the established link between economic
activity associated with a given level of warming
flows and emissions – and this corresponds to
(i.e., the damages).
FIGURE 3
D.CLIMATE framework
Damage
Model DAE-RGEM MAGICC D.CLIMATE
functions
Economic Temperature Economic Damaged
Output
growth and pathway impacts to growth path
emissions factors
pathway
Aligned with 3°C warming Reduced GDP 5%
Example
SSP2-6.0 at 2100 labour lower at
productivity 2070
With no change, Increased emissions Average temperature Climate change damage
economic growth result in a change change causes the impacts how land is used,
produces more in average tempertaure climate to change. This how people work and
greenhouse gas for different regions. results in physical how money is spent in
emissions globally. damages to the the economy. This
environment and world negatively impacts
around us. economic growth.
Note: The The
Note: temperature
temperaturepathway provided
pathway by MAGICC
provided is global
by MAGICC mean
is global surface
mean temperature.
surface TheThe
temperature. damage
damagefunctions e
stimate
functions
regional impacts
estimate based
regional on thisbased
impacts temperature pathway as well
on this temperature as other
pathway regional
as well climate
as other variables
regional (e.g.
climate precipitation), which
variables
are not
(e.g.provided by MAGICC.
precipitation), which are not provided by MAGICC.
Source:
Sources: Deloitte
Deloitte Economics
Economics Institute,
Institute, Meinshausen
Meinshausen et et
al. al (2011,
(2011, 2020),
2020), Nicholls
Nicholls et et
al.al(2021).
(2021).
7Europe’s turning point
Translating this concept into a modelling process Economic Outlook database that provides historical
involves three models that are linked through three and forecast GDP growth over the period 1980 to
key outputs. The Deloitte Economics Institute’s 2025.15 These growth rates are extrapolated using
approach extends methods adopted by the historical growth rates and assuming a degree of
Australian Bureau of Agricultural and Resource convergence over the long run.
Economics and Sciences (ABARES), the IPCC and
Population growth rates are calibrated using
other research organisations. The method is
the total population trajectories of the second
extended by necessity for practical public policy
socioeconomic pathway (SSP2) made available by the
purposes and the modelling is regionalised – allowing
International Institute for Applied Systems Analysis
results and insights to be produced at the regional
(IIASA). A linear interpolation is applied to build
level (such as countries or subnational regions).
yearly data.16 Labour supply is calibrated employing
The modelling process is summarised as follows: a similar approach and is assumed to broadly reflect
trends in population growth.
1. Deloitte’s in-house regional general equilibrium
model (DAE-RGEM) is used to produce a Unemployment rates are calibrated using short-term
projected path for economic output and forecasts developed by the IMF17 and extrapolated
emissions that align with a chosen SSP and RCP. using a moving average. This approach implicitly
assumes a steady state unemployment rate over the
2. For each RCP scenario the associated climate
medium to long term.
data (global mean annual surface temperature
increases and atmospheric concentrations) is 2.2.2 Emissions, energy efficiency and
sourced from a climate change model – MAGICC productivity improvements
version 7 (MAGICC7). 13 Separately, regional
In the base year, once-off shocks are used to
average temperature, precipitation and relative
calibrate the energy mix for each region to ensure an
humidity variables are sourced from a synthesis
accurate reflection of the current state of the energy
of models available from the Sixth Coupled
mix between renewable and traditional sources.
Model Intercomparison Project (CMIP6).14
These shocks are calibrated drawing on data from
3. This climate data is then fed into damage Our World in Data.18
functions to inform how shifts in temperature
The emissions trajectory for the baseline is
may play out in terms of impacts on the stocks
calibrated to align with the RCP6.0 emissions
and productivities of factors of production in
scenario, developed by the IPCC. RCP6.0 is chosen as
each sector/region. Unlike most other models,
an intermediate baseline scenario as it includes no
we model a broad range of damages, including
specific or significant policy effort to mitigate, acting
capital damages, sea-level rise (SLR) damages
as an appropriate baseline for reference. Emissions
to land stock, heat stress damages on labour
are calibrated via uniform shocks to emissions
productivity, human health damages to labour
efficiencies for all regions.19
productivity, agricultural damages from
changes in crop yields and tourism damages to In addition to these specific calibrations, a uniform
net inflow of foreign currency. energy efficiency improvement (0.5 per cent per
annum) is applied across all regions, reflecting a
2.2 Baseline economic assumptions
continuation of the long-run improvement that has
In the baseline, a set of assumptions have been been observed to date.
applied for macroeconomic growth rates and
2.3 Database: Regions and sectors
technological improvements over the period 2015
to 2070. These key variables have been calibrated The core economic data underpinning DAE-RGEM
drawing on historical and forecast time series from a – the social account matrix (SAM) – is sourced from
range of reputable sources. the GTAP database (Walmsley et al., 2013).20 This
economic data is supplemented with specific data
2.2.1 Macroeconomic variables
on electricity differentiated by power generation
Macroeconomic variables including GDP, population type (i.e. coal, gas, solar, etc.) from the GTAP satellite
and labour supply, and unemployment rate are database GTAP-Power as well as CO2 and non-CO2
calibrated for each year over the model period emissions data.21 The behavioural parameters are
to 2070. also sourced from GTAP for the most part with some
exceptions as discussed below.
Growth rates for GDP are calibrated drawing on data
from the International Monetary Fund’s (IMF’s) World This data is transformed in two key processes.
8Technical appendix
2.3.1 Regional aggregation region was isolated in the model with several
individual countries (the UK, France, Germany and
D.CLIMATE is a global model and can be tailored
Italy) and regional aggregations modelled within this
to a specified regional concordance in line with the
geographical area. The regional concordances for
GTAP database.22 For this project, the European
this study are presented in figure 4.
FIGURE 4
Regional concordance
Subregion name Country or area GTAP country code
Presented in Europe report
United Kingdom of Great Britain United Kingdom of Great Britain GBR
and Northern Ireland and Northern Ireland
France France FRA
Italy Italy ITA
Germany Germany DEU
Eastern Europe Belarus BLR
Bulgaria BGR
Czechia CZE
Hungary HUN
Poland POL
Moldova XEE
Romania ROU
Russian Federation RUS
Slovakia SVK
Ukraine UKR
Northern Europe Åland Islands FIN
Guernsey XER
Jersey XER
Denmark DNK
Estonia EST
Faroe Islands XER
Finland FIN
Iceland XEF
Ireland IRL
Isle of Man XER
Latvia LVA
Lithuania LTU
Norway NOR
Svalbard and Jan Mayen Islands NOR
Sweden SWE
Southern Europe Albania ALB
Andorra XER
Bosnia and Herzegovina XER
Croatia HRV
Gibraltar XER
Greece GRC
9Europe’s turning point
Subregion name Country or area GTAP country code
Presented in Europe report
Holy See XER
Malta MLT
Montenegro XER
North Macedonia XER
Portugal PRT
San Marino XER
Serbia XER
Slovenia SVN
Spain ESP
Western Europe Austria AUT
Belgium BEL
Liechtenstein XEF
Luxembourg LUX
Monaco XER
Netherlands NLD
Switzerland CHE
Source: Deloitte Economics Institute analysis of GTAP database.
2.3.2 Sectoral aggregation for the European region. However, there was a
specific effort made to distinguish two non-GTAP
D.CLIMATE can also be tailored to a specified
sectors (hydrogen and bioenergy) to aid in the
sectoral concordance in line with the GTAP
representation of the transition to net zero.
database.23 For this project, a relatively high-
level sectoral aggregation was chosen given The sectoral concordance for this
the level of regional detail that was required study are presented in figure 5.
FIGURE 5
Sectoral concordance
Abbreviation Sector name GTAP sector(s)
AGRI Agriculture, forestry, fishing Paddy rice
Wheat
Cereal grains
Vegetables, fruit, nuts
Oil seeds
Sugar cane, sugar beet
Plant-based fibres
Crops
Bovine cattle, sheep and goats,
horses
Animal products
Raw milk
Wool, silk-worm cocoons
Fishing
FORESTRY Forestry Forestry
10Technical appendix
Abbreviation Sector name GTAP sector(s)
COAL Coal Coal
OIL Oil Oil
GAS Gas Gas
OMIN Other mining Other mining
FOODMAN Food manufacturing Bovine meat products
Meat products
Vegetable oils, fats
Dairy products
Processed rice
Sugar
Food products
Beverages, tobacco products
LIGHTMAN Light manufacturing Textiles
Wearing apparel
Leather products
Wood products
Paper products, publishing
HYD Hydrogen Petroleum, coal products*
BIO Bioenergy (carbon-neutral) Petroleum, coal products*
P_C Petroleum, coal products Petroleum, coal products
HEAVYMAN Heavy manufacturing Chemical products
Basic pharmaceutical products
Rubber and plastic products
Mineral products
Ferrous metals
Metals
Metal products
Computer, electronic and optical
products
Electrical equipment
Machinery, equipment
Motor vehicles and parts
Transport equipment
Other manufactured goods (e.g.
furniture)
ELYTND Electricity transmission and Electricity transmission, distribution
distribution
ELYDIRTY Fossil fuels Coal base load
Gas base load
Oil base load
Other base load
Gas peak load
Oil peak load
ELYCLEAN New energy sector Nuclear base load
11Europe’s turning point
Abbreviation Sector name GTAP sector(s)
Wind base load
Hydro base load
Hydro peak load
Solar peak load
GDT Gas manufacture and distribution Gas manufacture, distribution
WATER Water Water
CONS Construction Construction
TRADE Retail trade and tourism Trade
Accommodation, food and service
activities
TRANS Transport Transport
Water transport
Air transport
Warehousing and support activities
OSERV Other services Communication
Financial services
Insurance
Real estate activities
Business services
Recreational and other services
Dwellings
GOVSERV Government services Public administration and defense
Education
Human health and social work
activities
*
The hydrogen and bioenergy sectors are not identified as individual sectors in the GTAP database but have instead been
distinctly separated from the petroleum, coal products sector. An explanation of this process is provided in the following
section.
Source: Deloitte Economics Institute analysis of GTAP database.
2.3.3 Commodity splits the split executed so as to maintain the following
high-level facts:
In an effort to provide greater granularity in the
representation of the transition to net zero, the y The size of the hydrogen sector is approximately
hydrogen and bioenergy sectors were split from 2 per cent of the parent sector (petroleum, coal
the parent sector: petroleum, coal products. This products). Its cost structure is different in that
process was required as the GTAP database does it draws more heavily on coal and P_C (i.e. the
not specifically identify either of these emerging parent sector itself) although there is sufficient
energy sectors individually. flexibility in its production function to allow for
a shift towards production using zero-emission
The petroleum, coal products sector was targeted
electricity and primary factors as the main
as the parent sector due to the similarities in its
inputs. The sales structure is the same as
sales structure to that of hydrogen and bioenergy.
its parent.
This transformation was informed by information
gathered on the current size of the hydrogen, y The size of the bioenergy sector is
bioenergy and petroleum, coking sectors and approximately 1.4 per cent of the parent sector
the respective cost and sales structures of each (petroleum, coal products). It relies solely on
individual sector. This research was gathered and the output of agriculture and waste as inputs to
12Technical appendix
production in conjunction with primary factors. Neither RCPs nor SSPs are ‘complete’ without the
The sales structure is the same as its parent. other. RCPs generate climate projections that do
This is a subset of the broader bioenergy sector not correspond to specific societal pathways. SSPs
as it is exclusively carbon-neutral varieties. provide alternative societal futures, where climate
change impacts and policies are not present. Thus,
y The remaining P_C sector is essentially the same
it is increasingly common to find research, including
as the original GTAP sector, but slightly smaller.
the IPCC’s Sixth assessment report, adopting an
There is scope for further refinement of this process, integrated SSP-RCP scenario framework.27
drawing on more detailed data to help get a better
Following this SSP-RCP framework, the baseline
picture of production, consumption and export,
scenario for the modelling used in the present paper
specifically at the detailed regional level.
is broadly consistent with SSP2-6.0. It has been
2.4 Physical climate modelling for D.CLIMATE applied in over 150 studies between 2014 and 2019
and is one of the more commonly implemented
The future of climate change is inherently uncertain.
scenarios that reflect continued emissions growth
The rate at which CO2 and other pollutants
and temperature increase from today.28
accumulate in the earth’s atmosphere could follow
any number of trajectories, with each leading to Data consistent with the SSP2 narrative and
a wide range of physical climate effects varying in RCP6.0 climate scenario was integrated in
both scope and scale. What is certain, however, is D.CLIMATE, representing the baseline state. RCP6.0
that the average global temperature has been rising represents an economic future with a high rate
and will likely continue to rise until a sustained and of GHG emissions, where several technologies
concerted effort is made to decarbonise globally. and strategies are implemented to reduce GHG
emissions, and radiative forcing stabilises after 2100.
In 2011, a set of four emissions and warming
The economic and emissions profile consistent with
pathways were published to support consistent
RCP6.0 has the potential to result in an increase to
scenario analysis in the climate modelling
global average temperature in excess of 3°C.29 The
community.24 These so-called Representative
SSP2 narrative reflects a continuation of current
Concentration Pathways (RCPs) were selected as
social, economic and technological trends as well as
plausible future GHG emissions and atmospheric
slow global progress towards achieving sustainable
concentration trajectories extending out to 2100.
development goals.30
They are as follows:
2.4.1 Climate of global average temperature
y RCP2.6 (assumes stringent decarbonisation)
increase – MAGICC
y RCP4.5 and RCP6.0 (two central scenarios)
Emissions produced by Deloitte’s DAE-RGEM
y RCP8.5 (a high GHG emission scenario). model are translated into global mean surface air
The IPCC’s Fifth assessment report of 2014 adopted temperature (GSAT) relative to the preindustrial
these RCPs as core scenarios for long-term (1750) period based on these emissions
projections and assessments. trajectories using a reduced complexity climate
model. Specifically, the D.CLIMATE framework
In 2017, a concurrent research effort sought to utilises outputs from the MAGICC as described in
develop a similar set of consistent future scenarios Meinshausen et al. (2011) and Meinshausen et al.
for human development, the five so-called Shared (2020) and configured by Nicholls et al. (2021).31
Socioeconomic Pathways (SSPs).25 These include Global temperature increases are the main driver of
a range of societal factors such as demographics, climate impacts and are regionalised via the damage
human development (for example, health and functions. MAGICC does not provide regional
education), economic growth, inequality, governance, temperature outputs or regional climate impacts.
technological change and policy orientations.26 They
are as follows: 2.4.2 Other climate variables – CMIP6
y SSP1 – Sustainability Separately, regional average temperature,
precipitation and wet bulb globe temperature
y SSP2 – Middle of the road (WBGT) have also been used. The data for each
y SSP3 – Regional rivalry variable is the multimodel mean of 17 global climate
models (GCMs) for the modelled SSP-RCP future
y SSP4 – Inequality pathways, which are available from the CMIP6.32
y SSP5 – Fossil-fuelled development. The GCMs output was downloaded from the Earth
System Grid Federation portal and then processed
13Europe’s turning point
into monthly periods per geography/region across distribution of daily peak temperatures and relative
the modelled regions in Europe and the rest of the humidity. Altogether, this means that heat waves
world from present day to 2100. 33 are likely to become more frequent and increasingly
extreme for many countries.
Twenty-year averages of the GCM projections are
used here to assess the key signals for future climate When workers exert energy to perform physical
change across short- to long-term horizons. Each tasks, their bodies produce thermal energy and
20-year averaged period represents the climate of begin to heat up internally. For body temperature
the mid-year. For example, the average temperature to be maintained at a healthy level, thermal energy
projection for the period 2011 to 2030 is assumed to needs to be transferred to a cooler external
represent the climate in the 2020 horizon. environment. If body temperature exceeds 39°C, an
individual can suffer a heatstroke, and temperatures
2.4.3 Damage function overview
exceeding 40.6°C can be fatal. However, before
The fundamental ‘driver’ of economic damages is these serious health effects occur at lower levels of
rising temperature. As rising temperature induces heat exposure, workers can experience diminished
climate change, economic output (as measured by ‘work capacity’ or mental task ability, and increased
GDP) is impacted through the physical damages that accident risk.
affect productivity and/or the stock of production
To continue functioning at elevated body
factors (figure 6).
temperatures, workers can take instinctive actions
This study includes six regionalised damages to to reduce their work intensity or increase the
Europe: frequency of short breaks. This ‘slowing down’ of
activity (whether it occurs through self-instinct or
y heat stress damages to labour productivity
occupational health management interventions)
y human health damages to labour productivity results in reduced ‘work capacity’ and lower
y SLR damages to land and capital stock labour productivity.35
y chronic reductions in capital productivity This analysis estimates the effect of rising
temperatures and changing relative humidity levels
y agricultural damages from changes in crop yields on labour productivity using WBGT as a measure of
y tourism damages to net inflow of foreign heat stress. Analysis is conducted at a geography or
currency. regional level. It is assumed that changes in labour
productivity (an economic concept) are equal to
The following section outlines each damage and how changes in estimated work capacity (a physiological
they impact the economy. concept).
2.4.4 Heat stress damages on labour productivity The methodology follows an approach proposed by
A working environment that is ‘too hot’ can Kjellstrom et al. (2017). This approach utilises a series
negatively affect the health and safety of workers, as of functions describing the relationship between
well as restrict their ability to perform tasks and limit WBGT and labour productivity across three different
their productive capacity.34 For jobs where tasks are work intensities: 200W (equivalent to light manual
performed outdoors, it can be difficult for workers to labour, such as office work), 300W (equivalent to
moderate their heat exposure. The same can be moderate manual labour, such as manufacturing)
true for indoor jobs where air-conditioning is not and 400W (equivalent to high-intensity manual
readily accessible. labour, such as farming). Relationships have been
determined by Kjellstom et al. (2017), based on a
Climate change is expected to see average global review of epidemiological datasets.
temperatures continue to rise, leading to shifts in the
FIGURE 6
‘Two-stage’ economic damages relationship
Climate change
Impact on land,
Change in average damages
Stage 1 Stage 2 labour and capital
temperature (i.e. change in rainfall
(i.e. GDP)
patterns, sea-level rise etc.)
Source: Deloitte
Deloitte Economics
Economics Institute.
Institute.
14Technical appendix
Workers in each GTAP sector are assumed to Climate change would allow diseases to invade
perform tasks at one of the three work intensities immunologically naïve populations with unprepared
specified above. GTAP sectors have been allocated medical systems and would affect food- and
to specific work intensities based on internal advice waterborne diseases, with cholera and diarrhoea
from Deloitte subject matter experts. being potentially most problematic.41
Consistent with the approach proposed by As extreme weather events become more severe
Kjellstrom et al. (2017), it is assumed that a and frequent, so too does the threat they present
geography or region’s WBGT varies over three to human populations. Climate change can affect
4-hour intervals comprising the approximate 12 air quality, leading to greater incidence of diseases
hours in a working day: caused by air pollution – the 2020 summer of
bushfires in Australia are a stark reminder of this.
1. early morning and early evening – four hours
Climate change may also affect human health
at WBGT mean (calculated using average
indirectly, through changes in food production, water
monthly temperature)
resources, migration and economic development.42
2. middle of the day – four hours at WBGT max
Human health is therefore prominent in estimates
(calculated using average monthly maximum
of future climate change impacts. The welfare
temperature)
costs (or benefits) of health impacts contribute
3. hours in between – four hours at WBGT half substantially to the total costs of climate change.
(calculated as the mid-point between WBGT Many estimates of economic damages rely on direct
mean and WBGT max). costs methodologies (i.e. price times quantity). With
regard to human health, the price is typically equal
These three variants of WBGT have been projected
to the value of a statistical life, based on estimates
at monthly intervals using the simplified WBGT
of willingness to pay to reduce the risk of death or
index – sWBGT – based on surface temperature and
diseases, or the willingness to accept compensation
water vapour pressure (developed by the Australian
for increased risk.43 However, these methods ignore
Bureau of Meteorology).36 Water vapour pressure
the human health impacts on labour productivity
was derived using estimates of relative humidity and
and the demand for health services.
the corresponding surface temperature.
The approach adopted for this analysis is based on
Labour productivity is then estimated for each
the work undertaken by Roson and Sartori (2015),
geography/region at monthly intervals, across each
which in turn is based on Bosello et al. (2006), by
of the three 4-hour intervals assumed to comprise
considering some vector-borne diseases (malaria,
the working day. The mean of these three estimates
dengue, schistomiasis), heat- and cold-related
is then taken to represent the average labour
diseases, and diarrhoea. It does not consider other
productivity for workers throughout the working
diseases and impacts mentioned in the IPCC AR5
day. Workers are assumed to maintain the same
(2014), such as the effects of extreme events, heat
level of productivity for all days contained within
exposure effects on labour productivity (separately
each month. Monthly labour productivity estimates
considered), haemorrhagic fever with renal
are then averaged to give an aggregate measure of
syndrome, plague, chikungunya fever, Japanese
labour productivity for each year in the modelling
and tick-borne encephalitis, cholera and other
period.
(nondiarrhoea) enteric infections, air quality– and
2.4.5 Human health damages to labour nutrition-related diseases, allergic diseases and
productivity mental health.44
The impacts of climate change on human health are The starting point of the analysis presented in
many and complex.37 Increasing temperatures can Bosello et al. (2006) is a meta-analysis of the
increase heat-related health problems, particularly epidemiological, medical and interdisciplinary
for those with pre-established cardiovascular and literature to achieve the best estimates for the
respiratory disorders.38 Increasing temperatures additional number of extra cases of mortality
can also reduce cold-related health problems, and morbidity associated with a given increase in
again most prevalent in people with cardiovascular average temperature.45 The information obtained
disorders.39 in this research has been combined with data on
Climate change can impact the range, abundance the structure of the working population, to infer
and dispersion of species carrying diseases. Studies the number of lost working days. The changes in
generally agree that the prevalence of malaria morbidity and mortality are interpreted as changes
increases with increase in temperature. Other in labour productivity.
vector-borne diseases may increase or decrease.40
15Europe’s turning point
Roson and Sartori (2015) update the work of Bosello of LECZ used for agricultural production in each
et al. (2006) to account for recent literature on health geography is assumed to be equal to the proportion
impacts and studies mentioned in IPCC (2014), of total land area used for agriculture in that same
scaling up or down the variations in geography.
labour productivity.
This analysis extends the Roson and Sartori (2016)
The results of these studies are expressed as methodology to also capture urban land area
changes in average labour productivity for a 1°C lost due to SLR, again leveraging World Bank data
increase in temperature (implicitly assuming that describing the extent of urban area in LECZ. In
the relationship is approximately linear). For the low-lying and seacoast urban areas, residential and
purposes of this analysis, and to understand the commercial properties may incur physical damages
relationship between human health impacts, and economic activity that would otherwise occur
increase in average temperature and time, we in these urban areas will also need to transition to
regressed the variables to find an equation with a other geographies.
satisfactory fit for the relationship.
The process for estimating both components is
The analysis estimates the higher-order economic as follows:
effects (or indirect costs) of human health impacts
y The percentage of effective land area lost per
and variations in labour productivity. It is important
metre of SLR is calculated by multiplying the
to note that this methodology excludes induced
following factors – the percentage loss in coastal
demand for health care.
wetland (a proxy for loss of land due to SLR,
2.4.6 Sea-level rise damages to land and estimated using the HadCM3 climate model
capital stock under the A1b SRES scenario),47 the LECZ
area, the percentage of erodible coast and
As average global temperatures continue to rise,
relevant coastline.
land-based glaciers are melting, and water bodies
are experiencing thermal expansion. Together, these y Considering which proportion of total coast
factors cause the phenomenon of SLR. is suitable for agricultural (productive)/urban
purposes, the percentage of effective land
SLR can impact a geography’s total stock of land
change is adjusted by agricultural land area/
(an economic factor of production) through a
urban land area.
combination of erosion, inundation and salt
intrusion along the coastline. As the global stock of y The percentage change in agricultural and
land declines due to SLR, productive activity that urban land stock is computed by multiplying the
would otherwise occur on that land is also foregone. percentage of effective land change by metre of
SLR and the estimated SLR.
The extent of land lost to SLR will depend on several
geography-specific characteristics, including (i) the Notably, there are a number of limitations involved
composition of the shoreline (cliffs and rocky coasts with this methodology of measuring SLR. One is the
are less subject to erosion than sandy coasts and assumption that the area of LECZ in each region is
wetlands), (ii) the total length of the coastline, (iii) used for agricultural or urban purposes. This will not
the share of the coast that is suitable for productive be true of all regions, some of which have large areas
purposes (i.e. in agriculture or urban land) and (iv) of unproductive coastal areas.
the VLM.46
Further, the method used focuses primarily on
This report estimates land area lost due to SLR using the loss of total land stock. The method does not
a methodology proposed by Roson and Sartori explicitly consider damage to labour productivity
(2016), who estimated the mean SLR (in metres) due to forced displacement, and SLR damage to
associated with global mean surface temperature infrastructure, which is generally established to be
change from a series of regressions based on higher than damage to land stock.48 However, it is
data within the latest IPCC AR5 Report, while also widely assumed that submergence by SLR does
accounting for vertical land movement. not lead to damages to capital, because it is a slow
process and by the time SLR arrives the capital stock
The proportion of agricultural land lost per metre
will have fully depreciated and property markets will
of SLR is then estimated based on the findings of
have adjusted.49
Roson and Sartori (2016), as well as World Bank data
describing the extent of low-elevation coastal zones Further, as average rising increases (a gradual process),
(LECZ) for each geography or region. The proportion the impact of extreme or acute SLR (e.g. king tides
16Technical appendix
and storm surges) will cause damage that may not be This damage estimate does not measure the
fully reflected in this function.50 Financial and asset direct impact of any particular natural disaster
value shocks of SLR to coastal property will also not be to a particular location at a particular time in
captured through this damage, but may represent a future modelled periods. Nevertheless, the
significant economic risk if warming continues.51 implied relationship between temperature
change and capital damage in Forzieri et al. (2018)
2.4.7 Capital damages
indirectly and implicitly reflects the fact that, as
This study captures climate-induced capital damages global temperatures continue to increase above
as a function of global mean average surface preindustrial levels, the frequency and intensity
temperature. of natural hazards will rise in aggregate, and the
productivity of capital will fall on average.
Capital damages in this context are ‘measured as a
reduction in the capital productivity across sectors’, 2.4.8 Agricultural damages from changes in
that is, the output produced per unit of capital crop yields
input. Reducing the productivity of capital as a result
Climate change will see rising temperatures,
of climate change represents a need for firms to
higher concentrations of CO2 in the atmosphere
spend more on capital to achieve the same amount
and different regional patterns of precipitation.55
of output in every period. This effectively acts as
These factors all affect crop yields and agricultural
a depreciation rate and diverts investment from
productivity.
otherwise productive applications.
The effects of climate change on agricultural
Accounting for capital damages in this way
productivity are one of the most studied areas
represents a departure from existing economic
of climate change impacts. Yet, despite the many
impact modelling and integrated assessments of
existing studies and the extensive empirical
climate change. In some cases, capital damages
evidence, it is still difficult to identify some sort of
are included but at a highly aggregated level that
‘consensus’ for the impacts of climate change on
limits regional analysis. Often, reports discuss the
agricultural productivity. There are many factors at
exposure or risk of geographies to capital damages
play, including the role of adaptation behaviour by
but do not attempt to monetise an impact.
farmers, firms and organisations, including variety
The relationship between global mean average selection, crop rotation, sowing times, the amount
temperature and capital productivity is of fertilization due to higher CO2 concentration and
parameterised using projected data estimated by the actual level of water available for irrigation and
Forzieri et al. (2018).52 This data provides projections irrigation techniques.56
of estimated annual damages (€ millions)53 to capital
Modelling the economic consequences of yield
that would occur due to a climate change – induced
changes to understand the consequences of climate
increase in intensity and frequency of natural
change impacts on agriculture is important for two
hazards occurring in Europe.54
main reasons. Firstly, varying levels of agronomic and
The specific natural hazards captured in this study economic adaptation exist in the agricultural sector;
include heat and cold waves, riverine and coastal farmers can adjust how they grow a particular crop,
flooding, wildfires, subsidence and high wind the location and timing of crop growth will shift
speeds (excluding cyclones). The capital that is in response to climate change impacts; trade in
damaged by these natural hazards in this study is agricultural commodities will adjust; and consumers
described as ‘critical infrastructure’ including existing will be able to substitute goods as prices adjust.57
transport systems, renewable and nonrenewable Each of these adaptive responses will mediate the
energy generation plants, industries, water supply impacts of yield changes. Secondly, climate change
networks, roads, railways and education and health impacts will vary by crop and by region, changing the
infrastructures. The projections are available up to comparative advantage of countries, and creating
2080 and are provided at country and hazard levels. winners and losers in global agricultural markets.58
The relationship between temperature increase and The approach undertaken in this analysis provides
estimated annual damage in the analysis by Forzieri an estimate of productivity changes for the whole
et al. (2018) is used as a proxy for the annual change agricultural sector across the modelled regions.
in capital productivity in the D.CLIMATE model, The methodology is based on Mendelsohn and
by country and region. This effectively smooths Schlesinger (1999) and Cline (2007), where the
a stochastic process of natural disaster impacts variation in output per hectare is expressed as a
over time into an average annual damage estimate function of temperature, precipitation and CO2
captured by a reduction in capital productivity. concentration.59
17You can also read
