Investment, growth and development in South Africa

Annex A. Modelling and analytical frameworks

Overview of the modelling framework

The scenario analysis of the Current Policies and the Enhanced Policies scenarios is based on a modelling framework, primarily based on the OECD’s ENV-Linkages model (Château, Dellink and Lanzi, 2014[45]). ENV-Linkages is a dynamic, multi-sector, multi-region computable general equilibrium (CGE) model that links greenhouse gas emissions to economic activities. As a general equilibrium model, ENV‑ serves as an appropriate tool to compare different projections of the global economy under alternative climate policy developments. CGE models are however usually weak in evaluating short-term macroeconomic effects, such as the positive effects of investment on economic growth. To capture short- and medium-term investment effects, ENV-Linkages has been calibrated relying on simulations with the National Institute Global Econometric Model (NiGEM) model, developed by the National Institute of Economic and Social Research (NIESR) (Hantzsche, Lopresto and Young, 2018[46]). NiGEM is a structural macro-econometric model of the world economy, used for the OECD Economic Outlooks. NiGEM captures demand, supply, monetary and financial dynamics, reflecting the current macro-economic outlook. NiGEM is also endowed with a climate module, which allows for an analysis of the economic consequences of climate policies, related revenue recycling as well as the effects of energy-related investment. The simulation of the Enhanced Policies scenario with the NiGEM model provides additional inputs to the ENV-Linkages model about the effect of investment on potential and effective GDP, employment, government balance and current accounts. The ENV-Linkages model would otherwise not capture the Keynesian-like dynamics of the effect of investment on potential GDP and corresponding investment‑multiplier effects.

This report combines a set of modelling input and tool (Figure A A.1). Inputs from the IEA Global Energy and Climate model (IEA, 2023[48]) are used to calibrate scenarios in NiGEM and ENV‑Linkages. Specifically, energy-related investment, energy system projections and when available energy policies are used to calibrate the scenarios. The Current policies projection of the economic models is calibrated on the IEA Stated Policies scenario (STEPS) scenario while the Enhanced Policies scenario relies on the IEA Announced Pledges (APS) scenario. The OECD Long Term Model (LTM) (Guillemette and Turner, 2021[94]) (Guillemette and Château, 2023[47]) provides macro-economic projections to 2100 (including GDP, Investment, Employment, etc.) used to calibrate the ENV-Linkages model. While the LTM usually provides projections for 116 countries, the model was extended for this report to produce projections for 146 countries.

While not fully integrated in the economic modelling analysis, the emission and temperature pathways underlying the scenarios are used to quantify the economic benefits from reduced climate damages between the modelled scenarios. The methodology for calculating these benefits is also outlined in this Annex.

Box A A.1. Modelling Framework

The scenario analysis of the Current Policies and the Enhanced Policies scenarios is conducted using a modelling framework primarily based on simulations of the OECD’s ENV-Linkages model (Château, Dellink and Lanzi, 2014[45]). ENV-Linkages is a recursive-dynamic, multi-sector, multi-region computable general equilibrium (CGE) model that explicitly links greenhouse gas emissions to economic activities and is well suited to capture economy-wide effects of climate policies. Traditional computable general equilibrium (CGE) models, while well suited for assessing long-term structural and indirect effects, tend to understate near-term demand-side gains from climate-related investment. In pure CGE frameworks, additional investment in low-carbon technologies is often assumed to crowd out investment in other sectors, thereby dampening short-run growth effects. In contrast, neo-Keynesian models supported by recent empirical studies (Batini et al., 2022[95]; Hepburn et al., 2020[96]; Deleidi, Mazzucato and Semieniuk, 2020[97]) point to the existence of positive “green investment multipliers,” suggesting that well-targeted climate investments can stimulate aggregate demand and output in the short run while supporting the transition to a low-carbon economy. To capture the short-term effects of investment on economic growth, ENV-Linkages is complemented with the macroeconometric growth model NiGEM (National Institute Global Econometric Model (Hantzsche, Lopresto and Young, 2018[46])). The integration of ENV-Linkages and NiGEM enables the joint assessment of long-term structural effects and short- to medium-term macroeconomic dynamics, including the investment-driven demand responses often absent from standard CGE frameworks.

To ensure consistency with plausible long-term economic and demographic developments the baseline of the ENV-Linkages model was calibrated on the OECD Long-Term Model (LTM) (Guillemette and Château, 2023[47]), which provides macro-economic projections up to 2100, including GDP, population, investment and employment. To ensure that international market and technology developments (e.g. global fuel prices, clean technology costs and shifting trade patterns) are reflected in the analysis, both scenarios are calibrated using inputs from the International Energy Agency’s World Energy Outlook (IEA, 2023[48]) based on the IEA’s Global Energy and Climate model (GECM) (IEA, 2025[49]).

Figure A A.1. Modelling Framework

Source: OECD/UNDP (2025[1]).

ENV-Linkages

The OECD’s dynamic computable general equilibrium (CGE) model ENV-Linkages is used as the basis to estimate the effect of climate mitigation policies on economic output. ENV-Linkages is a multi-sectoral, multi-regional model that links economic activities to energy and greenhouse gas emissions. A more comprehensive model description is given in Chateau, Dellink and Lanzi (2014[45]).

The baseline scenario’s macroeconomic calibration is described by Fouré et al. (2020[98]) and its sectoral calibration by Château et al. (2020[99]) while recent baseline results are illustrated in OECD (2024[100]).

The model is based on the Social Accounting Matrices (SAM) contained within the GTAP 11 database (Aguiar et al., 2023[101]). This database describes bilateral trade patterns, production, consumption and intermediate use of commodities and services, including capital, labour and tax revenues and use. The base year of the SAM and of the model is 2017. The short‑term changes to the economy reflect short-term economic changes from international databases: the OECD Economics Department (OECD, 2024[102]) and the International Monetary Fund (2024[103]).

In ENV-Linkages, the link between GHG emissions to economic activity is calculated using the GTAP emissions database (Aguiar et al., 2023[101]). There are several drivers of GHG emissions by activities, covering: CO₂ (fossil) from burning fossil fuels for any economic activity and fugitive emissions; CO₂ (non‑fossil) from industrial process emissions and land-use, land-use change and forestry sector (LULUCF); CH4 from rice cultivation, livestock production (enteric fermentation and manure management), fugitive methane emissions from coal mining, crude oil extraction, natural gas and services (landfills and water sewage); N2O from crops (nitrogenous fertilisers), livestock (manure management), chemicals (non‑combustion industrial processes) and services (landfills); Fluorinated gases (SF6, PFCs and HFCs) from chemicals industry (foams, adipic acid, solvents), aluminium, magnesium and semiconductor production. Non-CO₂ GHG emissions are converted into CO₂ equivalents using standard Global Warming Potential over 100 years coefficients defined in the IPCC’s 5^th Assessment Report (Myhre et al., 2013[104]).

The ENV-Linkages aggregation is very flexible and tailored for each study. For this report, the world is divided in 12 isolated countries and 8 more aggregated regions. Economic activity within a region in ENV-Linkages is aggregated into 42 economic sectors, but for sake of clarity, the analysis in this chapter will present nine aggregate sectors: three energy sectors (fossil-fuel, fossil fuel-power electricity and non-fossil power) and six end‑use sectors. The aggregation of sectors facilitates the analysis of key parts of the economy that are particularly affected by climate policies, such as emission-intensive industries, energy, transport or agriculture.

Table A A.1. ENV-Linkages regional Aggregation

ENV-Linkages region	GTAP countries
Canada	Canada
Germany	Germany
Spain	Spain
France	France
Italy	Italy
Poland	Poland
United States of America	United States of America
Other OECD EU Countries	Austria, Belgium, Czech Republic, Denmark, Estonia, Finland, Greece, Hungary, Ireland, Latvia, Lithuania, Luxembourg, Netherlands, Portugal, Slovakia, Slovenia, Sweden
EU non-OECD countries	Cyprus, Malta, Bulgaria, Croatia, Romania
Other OECD countries	Australia, Chile, Colombia, Costa Rica, Iceland, Israel, Japan, Liechtenstein, Mexico, New Zealand, Norway, South Korea, Switzerland, United Kingdom
China	China, Hong Kong
South Africa	South Africa
Türkiye	Türkiye
Other Latin America	Rest of North America, Argentina, Plurinational Republic of Bolivia, Brazil, Ecuador, Paraguay, Peru, Uruguay, Venezuela, Rest of South America, Guatemala, Honduras, Nicaragua, Panama, El Salvador, Rest of Central America: Belize, Rest of Caribbean, Haiti, Dominican Republic, Jamaica, Puerto Rico, Trinidad and Tobago
Other Eurasia	Albania, Belarus, Ukraine, Rest of Eastern Europe: Republic of Moldova, Rest of Europe, Serbia, Kazakhstan, Kyrgyzstan, Tajikistan, Armenia, Azerbaijan, Georgia, Rest of Former Soviet Union: Turkmenistan, Uzbekistan
India	India
Other East Asia	Rest of Oceania, Mongolia, Chinese Taipei, Rest of East Asia: North Korea and Macao, Cambodia, Indonesia, Lao People's Democratic Republic, Malaysia, Philippines, Singapore, Thailand, Viet Nam, Rest of Southeast Asia: Myanmar and Timor-Leste, Brunei Darussalam, Bangladesh, Nepal, Pakistan, Sri Lanka, Rest of South Asia: Bhutan and Maldives, Afghanistan
Other Africa	Egypt, Morocco, Tunisia, Rest of North Africa, Algeria, Cameroon, Côte d'Ivoire, Ghana, Nigeria, Senegal, Benin, Burkina Faso, Guinea, Togo, Rest of Western Africa, Mali, Niger, Central African Republic, Chad, Equatorial Guinea, Gabon, Congo, South Central Africa, Democratic Republic of the Congo, Ethiopia, Kenya, Madagascar, Malawi, Mauritius, Mozambique, United Republic of Tanzania, Uganda, Zambia, Zimbabwe, Rwanda, Rest of Eastern Africa, Comoros, Sudan, Botswana, Namibia, Eswatini, Rest of South African Customs Union: Lesotho
Russia	Russian Federation
Other middle east countries	Bahrain, Islamic Republic of Iran, Kuwait, Oman, Qatar, Saudi Arabia, United Arab Emirates, Rest of Western Asia, Jordan, Iraq, Lebanon, Syrian Arab Republic, Yemen

Table A A.2. ENV-Linkages sector aggregation

Aggregate Sector	Included ENV-L activities
Agriculture and food	All Crops (cro-a), Livestock (lvs-a), Forestry (frs-a), Fisheries (fsh-a), Food products (fdp-a)
Services	Collective services (osg-a), Electricity transmission and distribution (etd-a), Water supply; sewerage; waste management and remediation activities (wts-a), Business services (osc-a)
Construction	Construction (cns-a)
Energy-intensive and trade-exposed industry (EITE)	Iron and Steel (i_s-a), Paper & Paper Products (ppp-a), Non-metallic minerals (nmm-a), Chemical products (crp-a), Non-ferrous metals (nfm-a)
Other Industry	Textiles (txt-a), Electronics (ele-a), Fabricated metal products (fmp-a), Wood products (fmp-a), Motor vehicles (mvh-a), Electrical equipment (eeq-a), Machinery and equipment n.e.s. (ome-a), Rubber and plastic products (rpp-a), Basic pharmaceuticals (bph-a), Minerals n.e.s. (OMN-a)
Transport	Water Transport (wtp-a), Air Transport (atp-a), Transport n.e.s.: Land transport and transport via pipelines (otp-a)
Renewable and nuclear electricity	Hydro power (hyd-a), Wind power(wnd-a), Solar power (sol-a), Other power (xel-a), Nuclear power (nuc-a)
Fossil fuel-based electricity	Coal powered electricity (clp-a), Oil powered electricity (olp-a), Gas powered electricity (gsp-a)
Fossil fuel extraction and distribution	Coal extraction (coa-a), Crude Oil extraction (oil-a), Nat. gas: extraction plus manufacture & distribution (gas-a), Petroleum and coal products (p_c-a)

Production in ENV Linkages is assumed to operate under cost minimisation with perfect markets and constant returns-to-scale technology. Each production sector is composed of one representative firm that combines labour, capital and intermediate inputs (including energy) to supply one commodity. This production decision reflects the firm’s objective to maximise its profit taking as given prices of these inputs. The production technology is specified as nested Constant Elasticity of Substitution (CES) production functions in a branching hierarchy. This structure is replicated for each output, while the parameterisation of the CES functions may differ across sectors. The model adopts a putty/semi-putty technology specification, where substitution possibilities among factors are assumed to be higher with new vintage capital than with old vintage capital. In the short run this ensures inertia in the economic system, with limited possibilities to substitute away from more expensive inputs, but in the longer run this implies a relatively smooth adjustment of quantities to price changes. Capital accumulation is modelled as in the traditional Solow/Swan neo classical growth model, where economic growth is assumed to stem from the combination of labour, capital accumulation and technological progress. GHG emissions in ENV-Linkages are modelled either as linked to the use of the inputs (e.g. fossil fuels, chemical products, etc.) in the different sectors, or for process emissions related to the amount of sectoral output produced.

Household consumption demand is the result of static maximisation behaviour which is formally implemented as an “Extended Linear Expenditure System”. A representative consumer in each region – who takes prices as given – optimally allocates disposal income among the full set of consumption commodities and savings. Saving is considered as a standard good in the utility function and does not rely on forward looking behaviour by the consumer. The government in each region collects various kinds of taxes to finance government expenditures. Assuming fixed public savings (or deficits), the government budget is balanced through the adjustment of the income tax on consumer income. In each period, investment net-of-economic depreciation is equal to the sum of government savings, consumer savings and net capital flows from abroad.

International trade is based on a set of regional bilateral flows. The model adopts the Armington specification, if domestic and imported products are not perfectly substitutable. Moreover, total imports are also imperfectly substitutable between regions of origin. Allocation of trade between partners then responds to relative prices at the equilibrium.

Market goods equilibria imply that, on the one side, the total production of any goods or services is equal to the demand addressed to domestic producers plus exports; and, on the other side, the total demand is allocated between the demands (both final and intermediary) by domestic producers and the import demand.

ENV-Linkages is fully homogeneous in prices and only relative prices matter. All prices are expressed relative to the numéraire of the price system that is arbitrarily chosen as the index of OECD manufacturing exports prices. Each region runs a current account balance, which is fixed in terms of the numéraire.

As ENV-Linkages is recursive-dynamic and does not incorporate forward-looking behaviour, price‑induced changes in innovation patterns are not represented in the model. The model does, however, entail technological progress through an annual adjustment of the various productivity parameters, including e.g. autonomous energy efficiency and labour productivity improvements. Furthermore, as production with new capital has a relatively large degree of flexibility in choice of inputs, existing technologies can diffuse to other firms. Thus, within the CGE framework, firms choose the least-cost combination of inputs, given the existing state of technology. The capital vintage structure also ensures that such flexibilities are larger in the long run than in the short run.

Calibration of the Enhanced Policies scenario on the IEA’s World Energy Outlook

ENV-Linkages is used to compare changes in the South African economy under the Current Policies and under Enhanced Policies scenarios by 2035. The Current Policies scenario reflects policies already in place or legislated, incorporating the existing ambition and implementation gap in current NDCs. Under this scenario, global greenhouse gas (GHG) emissions decline by 6% by 2035 compared to 2022, leading to an estimated 2.45°C of warming by the end of the century. The Enhanced Policies scenario assumes accelerated climate policies and investments able to achieve emission reductions in line with the “well-below 2°C” goal of the Paris Agreement. Under this scenario, global emissions are reduced by 27% by 2035, putting the world on a pathway consistent with limiting global warming to 1.7°C by the end of the century. All country groupings reduce their emissions in the Enhanced Policies scenario compared to the Current policies scenarios, but at different rates.

The calibration of the two scenarios relies on inputs from the International Energy Agency’s World Energy Outlook (IEA, 2023[48]) projections (respectively STEPS and APS scenarios), which provides information on future policy-induced changes in the energy system.

NiGEM

As described in Hantzsche, Lopresto and Young (2018[46]), NiGEM is a global macroeconomic model, used by policymakers and the private sector for economic forecasting, scenario building and stress testing. It consists of individual country models for the major economies that are linked through trade in goods and services and integrated capital markets.

The individual country models within NiGEM are economy-wide systems of dynamic equations, based around the internationally accepted national accounting framework, with parameters estimated from aggregate time-series data. They all have broadly the same New Keynesian structure in that agents in the model are assumed to have rational expectations, though this can be amended and there are nominal rigidities that slow the process of adjustment to external events. Importantly, the individual country models incorporate well-specified supply-side behaviour that underpins the sustainable growth rate of each economy in the medium term. As far as possible the same theoretical structure has been adopted for each of the individual country models, except where clear institutional or other factors prevent this. As a result, variation in the properties of each country model reflects genuine differences emerging from estimation, rather than different theoretical approaches.

All country models contain the determinants of domestic demand, export and import volumes, prices, current accounts and net assets. In the long run, output is tied down by factor inputs and technical progress interacting through production functions. In the short run, the dynamic properties of the model are consistent with the data and well-determined.

International linkages come from patterns of trade, the influence of trade prices on domestic prices, the impacts of exchange rates and patterns of asset holding and associated income flows. The structure of the trade block ensures overall global consistency of trade volumes by imposing that the growth of import volumes is equal to the growth of export volumes at the global level. Trade volumes and prices are linked by Armington matrices, based on 2016 trade patterns. The volumes of exports and imports of goods and services are determined by foreign or domestic demand, respectively and by competitiveness as measured by relative prices or relative costs.

The investment shock introduced in the NiGEM model comprises spending on the industry end-use sector, energy supply and power generation infrastructure investments but excludes additional investments into energy efficiency and transport and buildings end-use sectors. Infrastructure investments (e.g. electricity grids, industrial equipment) add to the economy’s productive capital stock expanding the potential output of the economy in the long run, while energy-efficiency investments are not assumed to add new productive capacity in the same way.

Projected damages and avoided damages under different damage functions

While not fully integrated in the economic modelling toolkit, the scenario analysis presented in Chapter 2 includes estimates of the economic benefits from reduced climate damages between the modelled scenarios. Benefits from reduced climate damages are quantified starting from the emission and temperature pathways underlying the Current Policies and Enhanced Policies scenarios. These scenarios are on pathways that would imply an average global temperature increase by the end of the century of 2.4°C and 1.7°C, respectively.

Considering the uncertainty in the quantification of economic costs of climate change, for each scenario, this report employs three different damage functions (Barrage and Nordhaus, 2024[56]; Howard and Sterner, 2017[57]; Bilal and Känzig, 2026[58]), aiming to reflect the broad range of estimates from the literature. While virtually all studies agree that the global long-term damages from climate change will be substantial, there is significant uncertainty surrounding the exact magnitude of these impacts. Although specific channels through which climate change can affect economic and welfare outcomes are empirically well-established for a range of channels and across multiple sectors, quantifying the aggregate expected economic impact of climate change on economic output has proven to be challenging and remains an active and heavily debated area of research (O’Neill, 2022[105]; Howard and Sterner, 2017[57]; Tol, 2024[106]).

Uncertainty surrounds each step needed for the calculations of the economic benefits from reduced climate damages. The climate system’s response to rising GHG concentrations, which forms the basis to calculate economic damages, depends on the extent to which rising GHG concentrations translate into temperature increases, sea-level rise or the likelihood and the frequency and severity of highly stochastic extreme weather events (O’Neill, 2022[105]). Similarly, there are uncertainties on how these climate projections translate into economic damages. Climate change affects socioeconomic systems through both gradual and sudden impacts and these effects are difficult to predict in both timing and location. Moreover, damages will unfold over decades, with cumulative effects that are not immediately observable. Researchers have employed a variety of methodological approaches to estimate the economic damages. These approaches differ in terms of their underlying assumptions, data sources, methods and model structures, leading to a wide range of estimates in the literature.

The three damage functions employed for the analysis in this report are representative of the much wider set of damage curves existing in the literature. The curves rely on different approaches and lead to diverging estimates. The 2023 DICE damage function (Barrage and Nordhaus, 2024[56]) represents a low-end estimate. The function is calibrated using an enumerative strategy and is widely used in integrated assessment models. As a middle-ground, the damage function from Howard and Sterner (2017[57]) is based on a meta-analysis of previous studies on the global impacts of climate change. At the higher end, a reduced form function calibrated around the findings of Bilal and Känzig (2026[58]) is employed. This function is based on econometric estimates of the effect of global temperature shocks on global GDP, using cross country data. The medium- and high-range damage curves were chosen to align to the OECD Long-Term Macroeconomic Model (LTM) (OECD, 2025[107]), which incorporates climate damages, also reflecting a range of estimates.

The three damage functions used in this report have the same functional form and express global climate damages (${wld\_damage}_t$) as the fraction of global output lost in year t relative to a no-warming scenario. The calculation of the global output loss relies on a reduced-form quadratic function of the global surface temperature anomaly ($T_t^S$) in year t compared to pre-industrial levels (1850-1900). For Howard and Sterner and DICE 2023 the single parameter in the global damage function is aligned with the author’s preferred specification from the respective publications. For the Bilal and Känzig damage function, the parameter from the most recent update of the OECD Long-Term Macroeconomic Model is employed (OECD, 2025[107]):

Howard and Sterner: ${wld\_damage}_t=0.010038$ $\bullet {\left(T_t^S\right)}^2$

DICE 2023: ${wld\_damage}_t=0.003467$ $\bullet {\left(T_t^S\right)}^2$

Bilal and Känzig: ${wld\_damage}_t=0.048$ $\bullet {\left(T_t^S\right)}^2$

As the modelling in this study only projects emissions until 2035, it is assumed that emission pathways consistent with the Enhanced Policies and Current Policies scenarios are maintained beyond 2035 and follow the ones of the IEA’s scenarios on which this report’s scenarios were calibrated to (IEA, 2023[48]).1 However, due to limited data availability for the IEA scenarios, temperature trajectories had to be approximated using the 2024 climate scenarios of the Network for Greening the Financial System (NGFS, 2024[59]). Specifically, temperature projections for the Enhanced Policies scenario are approximated using the median (50^th) percentile of the temperature distribution estimated with the Model for the Assessment of Greenhouse Gas Induced Climate Change (MAGICC) climate model NGFS “Below 2°C” pathway. Likewise, projections for the Current Policies scenario are approximated using the NGFS “Fragmented World” pathway. Both approximations are chosen based on the alignment of the end of century temperatures with the projected temperature levels for the corresponding IEA WEO scenario.2

The disaggregation of global economic damages into region-level estimates for the three damage functions is further aligned with the OECD LTM (OECD, 2025[107]).The attribution rule is designed to meet two broad objectives: 1) individual-country damages aggregate up to the global damage function and 2) the cross-country distribution of damages relate to country-specific characteristics in a way consistent with the scientific literature. Country-specific damage factors are based on a corresponding reduced-form estimation from a meta-analysis of Tol (2024[106]), which suggests that most of the cross-country variance in economic damages from climate change can be reduced into two dimensions: i) current climate conditions and ii) national income levels. Subnational differences in impacts can be significant, but such regional differences could not be considered.

Kaya decomposition methodology

Figure 1 decomposes South Africa's historical emissions into contributions from underlying economic and energy-system drivers, using the extended Kaya identity framework. The Kaya identity is a well-established framework for attributing changes in GHG emissions to their underlying drivers (Kaya and Yokobori, 1997[108]) which has been widely applied in the literature (Lamb et al., 2021[109]; Le Quéré et al., 2019[110]; Li et al., 2024[111]). The extended formulation used here decomposes changes in total GHG emissions into changes in population, GDP per capita, energy intensity of GDP, fossil share in energy, the emission intensity of fossil fuel energy and the intensity of the economy in non-energy-related GHG emissions. The emissions intensity of the total energy mix in this analysis is split up into two parts: The first is the share of fossil fuels in the energy mix, indicating the extent to which an economy relies on coal, oil or gas compared to renewable energy sources. The second part is the emissions intensity of fossil fuel energy, which captures the amount of GHG emitted per unit of energy obtained from fossil fuel combustion. Shifts towards less emission intensive fossil fuels, such as those from coal to gas, improvements in combustion technologies or the deployment of carbon capture and storage (CCS) technologies can all reduce this factor.

By quantifying the relative contribution of key socioeconomic and technological factors to recent emission trends, decomposition analysis helps to determine whether changes in emissions are primarily driven by changes in energy efficiency $\left(c.2\right)$, the energy mix $\left(c.3\right)$, or economic activity and population $(c.1)$. To capture a more comprehensive scope of GHG emissions, the conventional Kaya identity3 is extended by an additive term to also account for non-energy related emissions from agriculture, industrial processes and waste management. Total GHG emissions (excluding emissions from LULUCF)4 can thus be expressed as:

${GHG}_{Total}=GDP$ $\bullet \left(\frac{ {GHG}_e+{GHG}_{ne}}{GDP }\right)$ $=\underset{c.1}{\underbrace{\left(P\bullet \frac{ GDP}{P }\right)}}$ $\bullet [\underset{\left(c.2\right)}{\underbrace{\frac{E_{Total}}{GDP}}}$ $\bullet \underset{\left(c.3\right)}{\underbrace{\left(\frac{E_{FF}}{E_{Total}}\bullet \frac{{GHG}_e}{E_{FF}}\right)}}$ $+\frac{{GHG}_{ne}}{GDP}]\#$

Where:

• $\mathit{P}$ = Population

• $\frac{\mathit{ }\mathit{G}\mathit{D}\mathit{P}}{\mathit{P}\mathit{ }}$ = GDP per capita

• $\frac{{\mathit{E}}_{\mathit{T}\mathit{o}\mathit{t}\mathit{a}\mathit{l}}}{\mathit{G}\mathit{D}\mathit{P}}$ = Energy intensity of GDP (PJ/$)

• $\frac{{\mathit{E}}_{\mathit{F}\mathit{F}}}{{\mathit{E}}_{\mathit{T}\mathit{o}\mathit{t}\mathit{a}\mathit{l}}}$ = Share of fossil fuels in total energy supply

• $\frac{{\mathit{G}\mathit{H}\mathit{G}}_{\mathit{e}}}{{\mathit{E}}_{\mathit{F}\mathit{F}}}$ = Energy-related GHG intensity of fossil energy (i.e. emissions from fossil fuel combustion) (Gt CO₂e/ PJ)

• $\frac{{\mathit{G}\mathit{H}\mathit{G}}_{\mathit{n}\mathit{e}}}{\mathit{G}\mathit{D}\mathit{P}}$ = Non-energy-related GHG intensity of GDP (e.g. emissions from agriculture, industrial processes, fossil fuel exploitation and waste) (Gt CO₂e/ $)

Building on the extended Kaya Identity, the Logarithmic Mean Divisia Index (LMDI) decomposition approach is used to examine the main driving forces underlying the emissions trajectories in Figure 1. LMDI allows for a complete, consistent and residual-free decomposition of aggregate emission changes into contributions from each driver (Ang, 2005[112]). It allows to distinguish structural emission intensity reduction effects (technological effects e.g. through improvements in energy efficiency) from the impact of economic and demographic expansion. Figure 7, Panel B instead applies a log-growth-rate decomposition based on the same identity. Unlike the LMDI approach, which decomposes absolute changes in emissions, the log-growth-rate decomposition decomposes the annualised log growth rate of emissions into the annualised log growth rates of its drivers.

Notes

← 1.  The Current Policies scenario is calibrated on IEA’s Stated Policies scenario (STEPS) for and the Enhanced Policies scenario is calibrated on IEA’s Announced Pledges scenario (APS) (IEA, 2023[48]).

← 2.  Following the IEA (2023[48]) global mean temperature increase would be limited to approximately 1.7°C above pre-industrial levels by 2100 in the APS scenario. In contrast, under STEPS, temperature rise is projected to reach 2.4°C by 2100. These estimates correspond to the median (50^th) percentile of the temperature distributions derived using the Model for the Assessment of Greenhouse Gas Induced Climate Change (MAGICC). Similarly, the median temperature increases projected by the MAGICC climate model for the NGFS “Below 2°C” and “Fragmented World” scenarios (for which the underlying emission pathways are obtained through the REMIND-MagPIE integrated assessment model) is 1.75°C and 2.45°C, respectively (NGFS, 2024[59]).

← 3.  Conventional Kaya identity: $GHG = P$ $\times \frac{GDP}{P}$ $\times \frac{E}{GDP}$ $\times \frac{GHG}{E}$

← 4.  Total GHG emissions in the decomposition analysis exclude emissions and removals from Land Use, Land-Use Change and Forestry (LULUCF) due to large measurement uncertainties and the influence of non-economic factors such as climate variability, natural disturbances and land management cycles.