## wpiea2026157-source-pdf ## Source details **Canonical URL:** [wpiea2026157-source-pdf](https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026157-source-pdf.pdf) ## Other formats - [Markdown version](/-/media/files/publications/wp/2026/english/wpiea2026157-source-pdf.pdf.md) - [Structured JSON version](/-/media/files/publications/wp/2026/english/wpiea2026157-source-pdf.pdf.json) --- ### Executive Summary — Context and motivation - Global energy and material dynamics affect energy and resources security, economic growth, employment, balance of payments, and economic stability. - The world is not on track to meet Paris Agreement goals, and large subsidies prolong reliance on fossil fuels. - Paper contrasts two frameworks: the standard Energy Transition (ET) framework and the Generalized Jevons Paradox (GJP). ### Executive Summary — Standard Energy Transition (ET) framework: core assumptions - Assumes substitution dynamics govern historical and future dynamics (dominant energies and materials replace each other over time). - Assumes energy and material dynamics largely mirror technological dynamics, downplaying technological and material interdependence and path dependence. - Frames mitigation as substituting high-carbon energy sources with low-carbon ones, primarily via carbon pricing, without reducing global energy and material demand. ### Executive Summary — Empirical challenges to ET - Despite numerous technological innovations, total consumption of all energies and materials has increased over time (exceptions: sheep wool, asbestos, and mercury). - Evidence suggests material flows are governed by additive dynamics rather than substitution dynamics. - Building renewable electricity infrastructure generates a relatively small amount of CO2, but downstream uses powered by low-carbon electricity entail material consumption that depends on fossil fuels. - Rapid growth in emissions in the 20th century occurred with unprecedented economic and population growth; replication is unclear given current trends. ### Executive Summary — Key empirical findings - Documents large and growing embedded emissions and materials: energies and materials are deeply intertwined, interdependent, and governed by additive dynamics, leading to accumulation and entwinement of "clean" and "dirty" technologies and materials. - Finds a strong positive correlation between efficiency in resource use and total resource use at a global scale. ### Executive Summary — The Generalized Jevons Paradox (GJP) - GJP captures direct and indirect energy/material demand effects of long-term energy and material interdependencies (energy/material symbiosis) accompanying significant increases in resource use efficiency. - Emphasizes “stickiness” of global fossil and material consumption, even as some countries decarbonize electricity generation. - Extends Jevons Paradox by adding macroeconomic, institutional, and technological drivers that generate systemic rebound effects and material interdependencies. - GJP is an empirical observation to warrant caution, not an economic “law.” ### Executive Summary — Drivers reinforcing the GJP - Market and geoeconomic power. - Trade arrangements. - Financial factors. - Institutional drivers shape investment decisions that can entrench fossil fuels. ### Executive Summary — Limitations of technology-centered mitigation - Substitution and efficiency improvements reduce emissions only to a small extent and are insufficient to reach zero emissions. - Hard-to-abate sectors with no scalable zero-emission substitutes in the near/medium term: aviation, shipping, iron and steel, cement, plastics, fertilizers, agriculture, construction, and arms. - Persistent material interdependencies and new material demands put upward pressure on downstream consumption and emissions even when low-carbon materials exist. ### Executive Summary — Policy implications and recommendations - Exercise caution applying ET framework relying primarily on substitution and efficiency improvements. - Need demand-side policy frameworks to better allocate energy and material flows across sectors and countries. - Carbon pricing and technological diffusion alone may be inadequate; policies should address institutional and structural drivers (market power, trade, finance). ### Executive Summary — Research avenues - Detailed analysis of materials in supply chain vulnerabilities. - Political economy analysis of material flows. - Multidimensional welfare analysis of decarbonization scenarios. *IMF Working Paper — Executive Summary* ### Introduction — importance, environmental context, and emissions targets - Energy refers to primary, secondary and tertiary energy; materials are elements, constituents, or substances of which something is composed or can be made. - Energy accounts for over 75 percent of global emissions. - Seven of the nine identified planetary boundaries have been crossed. - To meet Paris objective of “well below” 2°C of warming: - global greenhouse gas (GHG) emissions need to decline by around 5 percent per year over the next decades - and by around 7 percent if one accounts for evidence of an ongoing sharp decline in carbon sinks. ### Introduction — critique of ET framework: four key implicit assumptions 1. Technological and material dynamics mirror each other, so deployment of low-carbon technologies induces commensurate reduction in high-carbon materials. 2. Relative prices between high- and low-carbon energies drive substitution. 3. Relatively high elasticity of substitution between “clean” and “dirty” technologies. 4. Increased energy efficiency does not drive, over time, an increase in demand for energy services (i.e., no macro-scale rebound). ### Introduction — limitations and material dynamics emphasis - ET overlooks entwinement of materials, absolute size of global economy, technological interdependencies, path dependence, and institutional drivers. - Material dynamics underrepresented in macroeconomics and climate-economics literatures. ### Introduction — GJP definition and core drivers - GJP: increased efficiency in resource use is associated with higher, not lower, emissions at the world-economy level. - Three key drivers of GJP: market and geoeconomic power; trade arrangements; international financial factors. - GJP is an empirical observation calling for caution in projecting future energy flows. ### Formal assumptions supported by historical evidence - Technological and material dynamics are not equivalent. - Elasticity of substitution between “clean” and “dirty” technologies tends to be relatively low (and sometimes zero). - Global rebound effect: without deep institutional change, increased energy efficiency tends to drive greater energy demand. ### Empirics objective - Assess whether data are consistent with: - (A) an energy transition (absolute decline in fossil fuel consumption), or - (B) an energy addition (increasing consumption of all energy sources, reflecting growing embedded emissions and materials). ### Present — key empirical findings (summary) - Absence of an energy transition in absolute terms: variations in shares mask increasing consumption levels for all energies. - Since 1800, consumption of all energies and materials has increased continuously. - Recent decades: rising renewables share accompanied fossil fuel consumption growth. - Increase in total, fossil and renewable energy consumption highest in the two decades after 2000. ### Annex I and Figure notes (data coverage and definitions) - Annex I analyzes mean, median and total energy and electricity mix for 117 and 120 countries over 1971-2023 and 2000-2023. - Renewables in panel 5 = hydropower, solar and wind; fossil fuels = coal, gas and oil. - Traditional biomass assumed constant since 2015 in some panels. ### Energy Addition — observations and global empirical relationships - Despite declining carbon intensity, emissions rose alongside increased energy efficiency and renewable deployment. - Two thirds of global energy investments are in renewables (IEA 2024). - As of November 2023, governments were planning to produce, in 2030, over twice the amount of fossil fuels consistent with the median 1.5°C pathway (UNEP 2023). - In 2024, China approved close to 67 GW in new coal-fired power capacity, and nearly 98 GW in new or resumed coal power projects. - Energy efficiency correlates positively with total resource footprints (carbon, water, nitrogen, electricity) since the 1960s-1970s. - Regression approaches (OLS, median quantile regression, Theil-Sen) confirm strong positive correlations. ### Statistical findings (selected) - Energy efficiency vs. total world GHG emissions (1900-2023): - R² = 0.88 (sample size of 69). - Least squares slope coefficient = 16.9. - Water use efficiency vs. total world water use (1960-2014): - R² = 0.84 (sample size of 52). - Nitrogen use efficiency vs. total world nitrogen use (1961-2023): - R² = 0.87 (sample size of 63). - Relationship described as U-shaped over 1960-2014. - Electrification: - Near-perfect correlation between share of electricity in global final energy consumption and both total energy consumption and total fossil fuel consumption over last 50 years. - R² values = 0.97 and 0.94 (sample size of 53). ### Embedded emissions, sectoral details, and hard-to-abate areas - Embedded emissions and materials in oil and gas extraction and refineries significant (1971-2023); trends accelerated in 2024 and 2025 due to electrification of oil/gas production and data center construction. - Energy is a complement in production; shocks have outsized economic impacts (example: German GDP impacts under gas supply reductions). - Hard-to-abate sectors: aviation, shipping, iron and steel, cement, plastics, fertilizers, agriculture, construction and arms. - Global food system emits 18 Gt CO2 equivalent (one third of global emissions): - Half of warming effect caused by deforestation and energy production. - Methane: 35 percent of food-system warming effect. - Nitrogen oxides (fertilizers): 11 percent. - Fluorinated gases (refrigeration): 4 percent. ### International trade, consumption-based emissions, and embeddedness - Many low-carbon power countries import carbon-intensive goods and services, resulting in high embedded emissions. - Trade-embedded shares (Tubiana et al. 2025): - Iron and steel: 34 percent. - Chemicals: 45 percent. - Agricultural products: 24 percent. - Aluminum: 50 percent. ### Constraints, persistence of fossil fuels, and systemic dynamics - Renewables can coexist with persistent fossil fuel capacity for grid stability and complementary uses. - In western China, 35 to 40 percent of wind and solar capacity recently built has been coupled with coal development. - By 2024, global emissions from the power sector were still rising despite boom in renewables. - Industrial embeddedness example: most steel (more than three-quarters) is produced with coal; oil production linked to large coal quantities (around one billion tonnes per year globally). ### Historical perspective on "energy transition" - Historical fuel-share shifts did not produce clean replacements or phase-outs; energy addition occurred. - Fuelwood, coal, and oil historically coexisted and absolute consumption rose. ### Energy Addition — historical metrics and technological sprawl - Key historical metrics: - 3.6 million cubic meters of firewood. - In the 1930s, around 7 tonnes of coal to make a car. - For every tonne of oil consumed in Britain, 2.5 tonnes of coal consumed (interwar calculation). - 300 meters of drill pipe in 1901; ten kilometers for Tiber Prospect in 2010. - 2 million OCTG in 1950; 7 million OCTG in 2010. - OCTG accounted for 69 percent of the steel used in US oil production (1967 report). - Steel accounted for 2.55 million tons out of 2.6 million in 1967. - Cement plants burn mainly coal: 70 percent (IEA) and 90 percent (cement manufacturers). - China consumes around half of the world's cement, half of which is used to build its infrastructures. - Technological sprawl example: - Electric motors estimated in the tens of billions as of 2025; production projections around one billion by 2023. - Electric motors accounted for around 45 percent of electricity consumption in 2011, generating around 6 Gt of CO2. ### Hypotheses about the future and risks to rapid emission declines - Material intensity of oil-dependent infrastructure implies considerable embedded emissions linked to coal consumption. - Sectoral trends: - Carbon intensity of steel production stagnated since the 2000s. - Carbon intensity of cement increased by 1.5 percent per year over the last decade. - Emissions from cement plants have tripled since 1990, accounting for 8 percent of global emissions. - Plastic responsible for 3 to 5 percent of global emissions; production quadrupled since 1990. - Nitrogen fertilizers: 1.5 percent of emissions at production stage; 5 percent including soil-transformed nitrous oxide. - Emissions associated with a low-carbon shift range from 70 to 395 GtCO2; average across scenarios 195 GtCO2. - Metal production generates approximately 10 percent of global emissions. - Committed/proposed infrastructure emissions: - Historically, existing fossil fuel energy infrastructure would cumulatively emit about 658 GtCO2; proposed power plants around 188 GtCO2. - Updated estimates as of 2022: existing and proposed infrastructures around 680 Gt and 129 Gt, respectively. - Transport and hybridization: - Worldwide sales of hybrid vehicles and EVs are almost as high as each other (at 10 and 12 percent of total global sales, respectively). - BYD in China now sells more hybrid cars than pure EVs. - Possible lower-carbon pathway: - Emissions appear to have plateaued over the past 18 months in China. ### Interpretation — technological sprawl, Jevons-style effects, and critiques - Efficiency gains enable technological sprawl, increasing overall material and energy demand. - Critique of mono-energy histories: focusing on single energy sources overlooks interactions with materials and embedded emissions. ### Box 1 — Expected profitability and investment in the solar industry (summary) - Solar industry has low barriers to entry and low expected profitability; business models and volatile spot prices depress profitability. - Oil and gas industry: high barriers to entry, oligopolistic, highly profitable. - Low-carbon technologies are more capital-expenditure intensive and more sensitive to cost of capital than fossil fuels. - Expected profitability plays a more important role than costs in determining renewable investment. - Supply concentration: top three producing countries account for about 65 percent, 50 percent, and 70 percent of global production of agriculture, energy, and minerals (IMF 2023, p. 74). - Military spending: - Rose from around US$1 trillion in 2000 to US$2.1 trillion in 2021. - Grew close to 30 percent between 2021 and 2024, reaching $US2.7 trillion in 2024. - EU’s ReArm Europe program implies an increase in military spending of around $US765 billion by 2029. - Central bank crisis-era asset purchases: preliminary estimates (Espagne et al. 2023) suggest implied global temperature rise of purchases made under crisis-era asset purchase programs of two major central banks during COVID-19 stood at 2.3°C to 3.1°C. ### Annex V — Formalization and Kaya-perspective implications - Aggregate production function with clean and dirty final-good sectors: - Yt = (Yct(ε−1)/ε + Ydt(ε−1)/ε)ε/(ε−1) - ε ∈ (0, + ∞) is elasticity of substitution between sectors. - GJP implies sectors are complements → ε < 1. - Meta-analyses on price elasticities: - Short-term price elasticity of energy demand: −0.21 (ranging from −0.29 to −0.02). - Long-term price elasticity of energy demand: −0.61 (ranging from −0.77 to −0.19). - Gasoline price elasticity between −0.6 and −0.3, implying ε < 1. - Conceptual contrast: - ET: substitution, high elasticity, risk paradigm, optimization. - GJP: energy symbiosis, additive dynamics, complementarity, uncertainty paradigm, precautionary principle. ### Annex V — Research priorities (three areas) 1. Detailed analysis of materials in supply chains and vulnerabilities (input-output analysis; industrial/green industrial policy design). 2. Political economy of material flows (coalitions, rent-seeking, market power, geoeconomics, trade, finance). 3. Multidimensional welfare analysis of scenarios (distributional implications, sector-specific decarbonization, non-monetary CO₂ utility). ### Conclusion — main messages and policy-relevant cautions - Main concept: GJP — higher efficiency in resource use at world-economy level leads to higher total energy/material use. - Historical pattern: additive dynamics and symbiosis; different energies accumulate rather than substitute fully. - Empirical evidence challenges two core ET assumptions: substitution dynamics and equivalence of technological and material change. - Policy implication: need to reduce energy demand and address institutional drivers (rent seeking and market power; geoeconomic pressures; international trade arrangements; financial flows). - A major economy (e.g., China) could drive sustained emission declines over decades, but this rests on strong assumptions. - Recommend adopting an uncertainty paradigm and policies able to shape institutional drivers of the GJP. ### Selected empirical regressions (preserved verbatim) - Energy efficiency and total emissions (1900-2023): - OLS coefficient on energy efficiency: 16.94*** (standard error 0.766), Constant -3.572* (1.889), N 68, R2 0.881. - Quantile regression coefficient: 16.13*** (0.887). - Theil-Sen slope: 17.265 (95% CI Lower 15.808, Upper 19.168); intercept -1.902. - Water use efficiency and total water use (1960-2014): - OLS coefficient: 198.6*** (12.42), Constant 1,006*** (134.9), N 52, R2 0.836. - Theil-Sen slope: 221.015 (95% CI Lower 187.555, Upper 259.193); intercept 1046.28. - Electricity share in final energy consumption and total energy/fossil fuel consumption (1971-2023): - Effect on total energy consumption (EJ): - OLS coefficient: 2,123*** (50.95), Constant -23.30*** (7.816), N 53, R2 0.971. - Theil-Sen slope: 2129.455 (95% CI Lower 1993.782, Upper 2238.717); intercept -39.242. - Effect on fossil fuel consumption (EJ): - OLS coefficient: 1,148*** (39.38), Constant 23.60*** (6.042), N 53, R2 0.943. - Theil-Sen slope: 1158.363 (95% CI Lower 1055.637, Upper 1241.488); intercept 11.418. ### Key statistics and figures (preserved verbatim) - 3.6 million cubic meters of firewood. - 7 tonnes of coal to make a car (1930s). - For every tonne of oil consumed in Britain, 2.5 tonnes of coal consumed (interwar calculation). - 300 meters of drill pipe in 1901; ten kilometers for Tiber Prospect in 2010. - 2 million OCTG in 1950; 7 million OCTG in 2010. - OCTG accounted for 69 percent of the steel used in US oil production (1967 report). - Steel accounted for 2.55 million tons out of 2.6 million in 1967. - Cement plants burn mainly coal: 70 percent (IEA) and 90 percent (cement manufacturers). - China consumes around half of the world's cement, half of which is used to build its infrastructures. - Carbon intensity of cement increased by 1.5 percent per year over the last decade. - Emissions from cement plants have tripled since 1990, accounting for 8 percent of global emissions. - Plastic responsible for 3 to 5 percent of global emissions; production quadrupled since 1990. - Nitrogen fertilizers: 1.5 percent of emissions at production stage; 5 percent including soil-transformed nitrous oxide. - Emissions associated with low-carbon shift range from 70 to 395 GtCO2; average 195 GtCO2. - Metal production generates approximately 10 percent of global emissions. - Committed emissions historically: 658 GtCO2 (existing) and 188 GtCO2 (proposed). - Updated (2022) committed emissions: around 680 Gt and 129 Gt, respectively. - Hybrid and EV sales: 10 and 12 percent of total global sales, respectively. - Emissions plateaued over the past 18 months in China. - Electric motors estimated in the tens of billions as of 2025; production projections around one billion by 2023. - Electric motors accounted for around 45 percent of electricity consumption in 2011, generating around 6 Gt of CO2 — twice as much as global emissions a century earlier. *IMF Working Paper — The Generalized Jevons Paradox and the Future of Energy (source PDF excerpts).* ### Executive Summary ...................................................................................................... ### Executive Summary ### Context and motivation - Global energy and material dynamics matter for: energy and resources security, economic growth, employment, balance of payments, and economic stability. - The world is not on track to meet Paris Agreement goals, and large subsidies prolong reliance on fossil fuels. - The paper contrasts two frameworks for thinking about the future of energy and materials: the standard Energy Transition (ET) framework and an alternative view summarized as the Generalized Jevons Paradox (GJP). ### Standard Energy Transition (ET) framework: core assumptions - Assumes historical and future dynamics are governed by substitution dynamics (dominant energies and materials replace each other over time, focusing on market shares). - Assumes energy and material dynamics largely mirror technological dynamics, implicitly downplaying technological and material interdependence and path dependence. - Frames mitigation as substituting high-carbon energy sources with low-carbon ones, primarily via carbon pricing policies, with no need to reduce global energy and material demand. ### Empirical challenges to the ET framework - Despite numerous technological innovations in the 20th and 21st centuries, total consumption of all energies and materials has increased over time (exceptions: sheep wool, asbestos, and mercury). - Evidence suggests material flows are governed by additive dynamics rather than substitution dynamics. - Building renewable electricity infrastructure generates a relatively small amount of CO2, but downstream uses powered by low-carbon electricity entail material consumption that depends on fossil fuels and will likely continue to do so (examples: heavy industry, transport, food consumption, production of biofuels leading to deforestation). - Rapid growth in emissions in the 20th century occurred with unprecedented economic and population growth; it is unclear whether similar increases are replicable given current demographic and growth trends. ### Key empirical findings - Documents large and growing embedded emissions and materials: energies and materials are deeply intertwined, interdependent, and governed by additive dynamics, leading to accumulation and entwinement of "clean" and "dirty" technologies and materials in the global economy. - Finds a strong positive correlation between efficiency in resource use and total resource use at a global scale. ### The Generalized Jevons Paradox (GJP): definition and interpretation - GJP captures direct and indirect energy/material demand effects of long-term energy and material interdependencies (energy/material symbiosis) that accompany significant increases in resource use efficiency. - Emphasizes the apparent “stickiness” of global fossil and material consumption, even as some countries decarbonize electricity generation. - Extends the classic Jevons Paradox by incorporating macroeconomic, institutional, and technological drivers that reinforce each other and generate systemic rebound effects and material interdependencies. - GJP is an empirical observation to warrant caution in projecting future energy flows, not an economic “law.” ### Drivers reinforcing the GJP - Market and geoeconomic power. - Trade arrangements. - Financial factors. - These institutional drivers shape investment decisions in ways that can further entrench fossil fuels in the global productive structure. ### Limitations of a technology-centered mitigation approach - Technological substitution and efficiency improvements reduce emissions only to a small extent and are insufficient to reach zero emissions. - Some sectors currently have no scalable substitution with zero-emission technologies in the near or medium term: aviation, shipping, iron and steel, cement, plastics, fertilizers, agriculture, construction, and arms. - Persistent material interdependencies and continuous generation of new material demands put upward pressure on downstream consumption of many materials, and thus on emissions, even when new low-carbon materials are available. ### Policy implications and recommendations - Call for caution in applying the ET framework that relies primarily on substitution and efficiency improvements. - Need for demand-side policy frameworks to better allocate energy and material flows across sectors and countries. - Recognize that carbon pricing and technological diffusion alone may be inadequate; policies should address institutional and structural drivers (market power, trade, finance) that influence investment and material flows. ### Research avenues and further analysis suggested - Detailed analysis of the role of materials in countries’ supply chain vulnerabilities. - Political economy analysis of material flows. - Multidimensional welfare analysis of decarbonization scenarios. *IMF Working Paper — Executive Summary* ### Introduction ### Introduction ### Importance of global energy and material dynamics - Energy and materials are crucial to the functioning of the global economy, affecting: - energy and resources security (e.g., dependence on imported fuels or other resources) - economic growth (energy and materials are key complements in production) - supply chain resilience - employment (in low-carbon energy and related sectors) - balance of payments (including by reducing energy import costs) - economic stability (e.g., via lower volatility induced by fossil fuel markets and other commodity markets) - Energy refers to primary, secondary and tertiary energy. Materials are elements, constituents, or substances of which something is composed or can be made. Dynamics refer to changes in quantities, prices, and global exchanges over time. ### Environmental impacts and the climate context - Energy accounts for over 75 percent of global emissions. - Climate change and biodiversity loss are compounding threats: - seven of the nine identified planetary boundaries have been crossed. - biodiversity decline affects ecosystem functioning, water availability and quality, food security and nutrition, and resilience to climate impacts. - The world is not on track to meet Paris Agreement goals; large subsidies prolong reliance on fossil fuels. - To meet the Paris objective of “well below” 2°C of warming: - global greenhouse gas (GHG) emissions need to decline by around 5 percent per year over the next decades - and by around 7 percent if one accounts for evidence of an ongoing sharp decline in carbon sinks. ### Emissions trends and historical context - Despite a steady decline in carbon intensity, emissions have continued to rise, driven by increased fossil fuel consumption. - Rapid growth in emissions in the 20th century occurred alongside unprecedented economic and population growth; it is unclear whether similar increases are replicable. - Absolute consumption of almost all materials has increased significantly in recent decades; no raw material has ever become obsolete. ### Energy security, balance of payments, and economic implications - Shifting from fossil fuels to renewable energy can: - bolster energy security by reducing dependence on imported fuels - create jobs in low-carbon energy and related sectors - enhance the balance of payments through lower energy import costs - support economic stability via lower volatility induced by fossil fuel markets (IMF 2025a, p. 26) - Many commodities-exporting countries depend on inflows of foreign exchange from fossil fuel exports to finance imports of consumption and investment goods, including those needed for low-carbon investments. ### Critique of the conventional Energy Transition (ET) framework - The conventional view: global energy and material dynamics are largely governed by substitution dynamics; relative price dynamics of energy services drive energy “transitions.” - Two versions of the ET framework: - Theoretical economics: strong assumptions on substitutability of low- and high-emission sources. - Technical systems approach: solid understanding of technical feasibility but uses models with largely/fully exogenous growth that do not capture systemic rebound effects. - Four key assumptions underpinning the ET framework (several implicit): 1. technological and material dynamics mirror each other, so deployment of low-carbon technologies induces commensurate reduction in high-carbon materials; 2. relative prices between high- and low-carbon energy technologies play a crucial role in substituting fossil fuels with low-carbon capital; 3. relatively high elasticity of substitution between “clean” and “dirty” technologies; 4. increased energy efficiency does not drive, over time, an increase in the demand for energy services (i.e., no macro-scale “rebound effect”). - The ET framework’s policy implications center on carbon pricing, subsidies, regulations, technological progress and diffusion, trade facilitation, and electrification. ### Limitations of the ET framework and emphasis on material dynamics - The ET framework overlooks or insufficiently accounts for: - entwinement of materials and resources in production - absolute size of the global economy - technological interdependencies and path dependence - broader environmental context (e.g., seven crossed planetary boundaries) - institutional drivers of material processes and investment dynamics - Key sectors create durable interdependencies between low-carbon technologies and fossil fuels: shipping, iron and steel, cement, plastics, fertilizers, construction, and arms. - Material dynamics have been underrepresented in macroeconomics and climate-economics literatures; standard production theory often omits explicit materials modeling. ### The Generalized Jevons Paradox (GJP) - Definition: increased efficiency in resource use (including energy consumption) is associated with higher, not lower, emissions. - The GJP extends the Jevons Paradox (original observation that higher energy efficiency ultimately leads to higher energy demand) to large, complex material-economic systems under institutional constraints. - Absent deep institutional change, the decarbonization potential of technological change that increases GHG or material efficiency is limited by effects on: - demand - trade - expected profits - induced technological developments - Three key drivers of the GJP: - market and geoeconomic power - trade arrangements - international financial factors - The GJP is an empirical observation calling for caution in projecting future energy flows, not an economic “law.” ### Formal assumptions supported by historical evidence - The history and functioning of the global energy system support three key assumptions: 1. technological and material dynamics are not equivalent; 2. the elasticity of substitution between “clean” and “dirty” technologies tends to be relatively low (and in some cases zero, given lack of available and/or scalable substitutes); 3. there is a rebound effect at the level of the global economy: without deep institutional changes, increased energy efficiency tends to drive greater energy demand. - Together, these imply that the global input-output matrix tends structurally to become larger over time, entrenching fossil fuels in the productive structure and producing “stickiness” in global emissions. ### Empirics: Energy Transition or Energy Addition? - Objective: examine global energy and material dynamics across three periods—present (recent decades), historical (19th and most of 20th century), and the future (hypotheses for next decades)—to assess whether data are consistent with: - (A) an energy transition (absolute decline in fossil fuel consumption), or - (B) an energy addition (increasing consumption of all energy sources, reflecting growing embedded emissions and materials). - Embedded emissions and materials: emissions and materials associated with a good (e.g., coal embedded in a tonne of steel; emissions embedded in manufacturing and shipping a car). - Present — key empirical findings: - Absence of an energy transition in absolute terms: variations in shares of global energy sources mask increasing consumption levels for all energies. - Since 1800, consumption of all energies and materials has increased continuously. - Recent decades: steady rise in the share of renewables in the energy mix has accompanied fossil fuel consumption growth. - In absolute terms, the increase in total, fossil and renewable energy consumption has never been as high as it was in the two decades after 2000. *Source: IMF Working Paper — Introduction* ### Annex I shows a similar picture by analyzing the change in the mean, median and total energy mix and electricity mix for ### The Generalized Jevons Paradox and the Future of Energy ### No Energy Transition, but an Energy Addition - Annex I analyzes changes in the mean, median and total energy mix and electricity mix for 117 countries and 120 countries over the period 1971-2023 and 2000-2023, respectively. - Key graphical evidence (Figure 1): - Global primary energy consumption from all energy sources (EJ per year): 1800-2024. - Global primary consumption of coal (EJ per year): 1800-2024. - Global energy consumption by source (percent of total primary energy mix): 1900-2024. - Global primary energy consumption (EJ per year): 1970-2024. - Share of renewables in primary energy consumption (percent) vs. total fossil fuel consumption (EJ): 1900-2024. - Increase in total energy and fossil energy consumption (EJ per twenty-year span): 1800-2019. - Notes from figure annotations: - In panels 1, 3 and 4, traditional biomass is assumed to be constant since 2015 given the absence of more recent data. - In panel 5, renewables correspond to hydropower, solar and wind; fossil fuels correspond to coal, gas and oil; the share of renewables is calculated as a percentage of total global primary energy consumption. - Panel 6 excludes the last twenty-year period past 2019 to exclude the impact of the COVID-19 pandemic on global energy. ### Energy Addition and the Generalized Jevons Paradox - Despite declining carbon intensity of the global economy, rising emissions have continued alongside increased energy efficiency and rapid renewable deployment. - Two thirds of global energy investments are in renewables (IEA 2024). - As of November 2023, governments were planning to produce, in 2030, over twice the amount of fossil fuels consistent with the median 1.5°C pathway (UNEP 2023). - In 2024, China approved close to 67 GW in new coal-fired power capacity, and nearly 98 GW in new or resumed coal power projects – the highest level since 2015 (Qi and Shearer 2025). - Global empirical relationships: - Energy efficiency (GDP per unit of resource use) correlates positively with total resource footprints for carbon, water, nitrogen, and electricity since the 1960s-1970s. - Regression approaches used: OLS, median quantile regression, and Theil-Sen estimator (Annex II) — all confirm strong positive correlations. - Specific statistical findings: - Energy efficiency (international dollars per tonne of CO2 equivalent) vs. total world GHG emissions (Gt CO2-eq) over 1900-2023: - R² = 0.88 (sample size of 69). - Least squares slope coefficient = 16.9 (interpreted: each dollar of increased energy efficiency associated with 16.9 Gt in additional emissions over 1900-2023). - Water use efficiency vs. total world water use over 1960-2014: - R² = 0.84 (sample size of 52). - Nitrogen use efficiency vs. total world nitrogen use for agriculture over 1961-2023: - R² = 0.87 (sample size of 63). - Relationship described as U-shaped over 1960-2014: efficiency fell 1960-1984, then increased 1985-2014; nitrogen use surged by a factor of five in the initial period, growing over three times faster than world GDP then, and growing only 11 percent over the next decade versus 33 percent for world GDP. - Electrification: - Near-perfect correlation between the share of electricity in global final energy consumption and both total energy consumption and total fossil fuel consumption over the last 50 years. - R² values = 0.97 and 0.94, respectively (sample size of 53). ### Embedded Emissions, Sectoral Details, and Hard-to-Abate Areas - Embedded emissions in oil and gas: - Electricity and gas consumed in oil and gas extraction and oil refineries show significant embedded emissions and materials (Figure 4, 1971-2023). - Trends accelerated in 2024 and 2025 due to electrification of oil and gas production and data center construction (Rystad Energy 2025). - Role of energy as complement in production: - Energy shortages/shocks have outsized economic impacts where energy is an input across sectors (e.g., chemicals, metallurgy). - Example: assuming a 30 percent reduction in gas supply, some models predicted GDP impacts for Germany of around 0.2-0.3 percent of GDP up to 3 percent in severe scenarios (Bachmann et al. 2024); actual German GDP at end-2023 was 6.5 percent lower than projections at end-2021 even though gas supply reduction was around 15-20 percent (Geerolf 2025). - Lifecycle and input-output perspectives: - Input-output matrices reveal entwinement of “green” and “brown” sectors; differences across goods/services in lifecycle emissions and energy symbiosis. - Hard-to-abate sectors and food system: - Partial decarbonization has mainly affected electricity (which accounts for 40 percent of global emissions), heating and land transport to a lesser extent. - Hard-to-abate sectors include aviation, shipping, iron and steel, cement, plastics, fertilizers, agriculture, construction and arms — technological prospects remain highly uncertain. - Global food system emits 18 Gt of CO2 equivalent (one third of global emissions): - Half of warming effect caused by deforestation and energy production. - Methane: 35 percent of food-system warming effect. - Nitrogen oxides (fertilizers): 11 percent. - Fluorinated gases (refrigeration): 4 percent. ### International Trade, Consumption-Based Emissions, and Embeddedness - Many low-carbon power countries import carbon-intensive goods and services, revealing high embedded emissions. - Figures show emissions embodied in international trade for France, the UK, and Norway (various time series 1970s-2024). - Empirical estimates of trade-embedded shares (Tubiana et al. 2025): - Iron and steel: 34 percent of emissions embedded in trade. - Chemicals: 45 percent. - Agricultural products: 24 percent. - Aluminum: 50 percent. - Broader evidence indicates significant resource use and biodiversity pressures embedded in trade and international value chains. ### Constraints, Persistence of Fossil Fuels, and Systemic Dynamics - Renewables deployment can coexist with persistent fossil fuel capacity due to grid stability, geographic patterns of demand, and complementary uses: - In western China, 35 to 40 percent of wind and solar capacity recently built has been coupled with coal development (Zhang 2023). - Chinese COP28 position: eliminating fossil fuels used to maintain grid stability described as “unrealistic”; preference for “transition away” rather than “phasing out” (Stanway 2023). - In the U.S., electricity demand growth over 3.5 percent per year driven by electrification of oil and gas production and data center construction (Rystad Energy 2025); electricity for data centers is mostly supplied by fossil fuels (Murthaugh 2025). - By 2024, global emissions from the power sector were still rising despite boom in renewables. - Industrial embeddedness: - Oil production and use are materially and energetically linked to steel production; most steel (more than three-quarters) is produced with coal, involving large coal quantities for oil production (around one billion tonnes per year globally). ### Historical Perspective on "Energy Transition" - Historical shifts in relative shares of fuels did not constitute a clean replacement or phasing out; instead, energy addition occurred: - Fuelwood consumption increased in most industrial countries in the 19th and 20th centuries. - Wood remained fundamental to energy systems (e.g., mining timber consumption in early twentieth-century British mines between 3 and 4.5 million cubic meters of pit-props per year). - The narrative of successive transitions (wood → coal → oil) overlooks absolute increases and coexistence of multiple fuels. *Source: IMF Working Papers — The Generalized Jevons Paradox and the Future of Energy.* ### 3.6 million cubic meters of firewood. ### wpiea2026157-source-pdf - 3.6 million cubic meters of firewood. ### Energy Addition - Historical evidence indicates expanding embedded emissions through a “symbiotic” relationship between oil and coal rather than a straightforward energy transition. - Key historical metrics: - In the 1930s, it took around 7 tonnes of coal to make a car. - An interwar calculation showed that for every tonne of oil consumed in Britain, 2.5 tonnes of coal had to be consumed to build cars and required infrastructure (Dalley 1935, p. 17). - In 1901, 300 meters of drill pipe were enough to bring oil to Spindletop in Texas; in 2010, the Tiber Prospect pierced the continental shelf ten kilometers below the surface of the ocean. - The United States consumed around 2 million OCTG in 1950 and 7 million OCTG in 2010 (U.S. Printing Office 1954, p. 1044). - A comprehensive 1967 report indicates OCTG accounted for 69 percent of the steel used in US oil production; steel accounted for the bulk of metal requirements (2.55 million tons out of 2.6 million in 1967) (National Petroleum Council 1967). - Cement and coal interdependence: - Worldwide, cement plants burn mainly coal (70 percent according to the IEA, 90 percent according to cement manufacturers) (IEA 2015). - China consumes around half of the world's cement, half of which is used to build its infrastructures. ### Hypotheses about the Future - Risks to rapid emission declines: - The material intensity of oil-dependent infrastructure implies considerable embedded emissions (steel, cement, etc.), which in turn are linked to coal consumption. - Sectoral trends limiting technological-driven emissions reductions: - Carbon intensity of steel production has stagnated since the 2000s. - Carbon intensity of cement has increased by 1.5 percent per year over the last decade. - Emissions from cement plants have tripled since 1990, accounting for 8 percent of global emissions. - Plastic is responsible for 3 to 5 percent of global emissions; production has quadrupled since 1990. - Nitrogen fertilizers are responsible for 1.5 percent of emissions at the production stage and 5 percent if soil-transformed nitrous oxide is included. - If low-carbon electricity powers a productive structure still dominated by large-scale and growing production and consumption of cars, cement, steel, plastic and industrial agriculture, global emissions will at best stagnate at a high level, and global warming will at best slow down but not stop. - Infrastructure for mitigation and its emission costs: - Emissions associated with a shift to a low-carbon energy system range from 70 to 395 GtCO2, with an average across scenarios of 195 GtCO2 (Slameršak et al. 2022). - Metal production generates approximately 10 percent of global emissions (Yokoi et al. 2022). - Building, operating, and maintaining mitigation-oriented infrastructure requires cement, steel, metals, glass and plastic—materials from hard-to-abate sectors. - Committed and proposed infrastructure emissions: - If operated as historically, existing fossil fuel energy infrastructure would cumulatively emit about 658 gigatonnes (Gt) of CO2; proposed power plants would emit around a further 188 Gt of CO2 (Tong et al. 2019). - Updated estimates as of 2022: existing and proposed energy infrastructures had risen to around 680 Gt and 129 Gt, respectively. - Energy variability and hybridization: - In low- and lower-middle-income countries, microgrids combining solar panels, hydropower, batteries and often diesel generators are growing; the number of diesel generators in these mini-grids continued to grow between 2018 and 2024. - In transport, hybrid vehicles remain significant: worldwide sales of hybrid vehicles and EVs are almost as high as each other (at 10 and 12 percent of total global sales, respectively). BYD in China now sells more hybrid cars than pure EVs. - Possible lower-carbon pathways and limits: - Emissions appear to have plateaued over the past 18 months in China (Myllyvirta 2025). - China’s rapid, large-scale deployment of low-carbon technologies—if it becomes the global economy’s dominant pole—could induce a lower-carbon global growth regime, though this would unfold over several decades. - China’s carbon intensity currently starts from a very high level — roughly on par with that of Saudi Arabia. ### Interpretation — Past, Present, and Material Dynamics - Past quantity vs. relative dynamics: - In volume terms, the UK consumed more wood for energy between 1900 and 1950 than in the 1750s; Warde (2007) places 3.6 million cubic meters as the upper range of peak fuelwood consumption in England and Wales in the mid-18th century. - Energy history’s transition-centric narratives (muscle/wood/water → coal/steam → oil/electricity/nuclear) emphasize relative shifts and understate absolute quantity dynamics and coexisting energy uses. - Technological sprawl and Jevons-style effects: - Efficiency gains enable wider application of technologies (“technological sprawl”), increasing overall material and energy demand. - Electric motors diffusion example: - Estimated in the tens of billions as of 2025. - Some projections put the production of new electric motors at around one billion by 2023 (Associated Press 2018). - Electric motors accounted for around 45 percent of electricity consumption in 2011 (Waide and Brunner 2011), generating around 6 Gt of CO2 — twice as much as global emissions a century earlier when manufacturers used steam engines. - Critique of mono-energy approaches: - Histories focusing on a single energy source (coal, oil, wood, or renewables) overlook interactions with materials (e.g., timber for coal extraction; concrete and steel for oil), leading to incomplete accounts of embedded emissions and material dependencies. ### Key statistics and figures (preserved verbatim) - 3.6 million cubic meters of firewood. - 7 tonnes of coal to make a car (1930s). - For every tonne of oil consumed in Britain, 2.5 tonnes of coal consumed (interwar calculation). - 300 meters of drill pipe in 1901; ten kilometers for Tiber Prospect in 2010. - 2 million OCTG in 1950; 7 million OCTG in 2010. - OCTG accounted for 69 percent of the steel used in US oil production (1967 report). - Steel accounted for 2.55 million tons out of 2.6 million in 1967. - Cement plants burn mainly coal: 70 percent (IEA) and 90 percent (cement manufacturers). - China consumes around half of the world's cement, half of which is used to build its infrastructures. - Carbon intensity of cement increased by 1.5 percent per year over the last decade. - Emissions from cement plants have tripled since 1990, accounting for 8 percent of global emissions. - Plastic responsible for 3 to 5 percent of global emissions; production quadrupled since 1990. - Nitrogen fertilizers: 1.5 percent of emissions at production stage; 5 percent including soil-transformed nitrous oxide. - Emissions associated with low-carbon shift range from 70 to 395 GtCO2; average 195 GtCO2. - Metal production generates approximately 10 percent of global emissions. - Committed emissions historically: 658 GtCO2 (existing) and 188 GtCO2 (proposed). - Updated (2022) committed emissions: around 680 Gt and 129 Gt, respectively. - Hybrid and EV sales: 10 and 12 percent of total global sales, respectively. - Emissions plateaued over the past 18 months in China. - Electric motors estimated in the tens of billions as of 2025; production projections around one billion by 2023. - Electric motors accounted for around 45 percent of electricity consumption in 2011, generating around 6 Gt of CO2 — twice as much as global emissions a century earlier. *IMF WORKING PAPERS — The Generalized Jevons Paradox and the Future of Energy (excerpts from source PDF).* ### 2. Embedded Emissions and Materials in 1930. ### 2. Embedded Emissions and Materials in 1930. ### Historical dynamics and material embeddedness - Wood consumption soared not in spite of coal, but because of it. - In 1900, railways in the U.S. consumed around 10 percent of domestic lumber. - In the 1950s, in Europe, between 2 and 4 cubic meters of wood were needed per 100 tonnes of coal. - In the 1930s, around 7 tonnes of coal were needed to produce a car. - Industrialization is better understood as a process of amplification and symbiotic expansion of all materials, i.e., a deepening embeddedness of emissions and materials in the global economy. ### Origins and evolution of the “energy transition” concept - Until the 1970s, the term “energy transition” was hardly used; experts expected shifts in composition and possible stabilization, not disappearance of entire energy sources. - Early conservation concerns prioritized coal depletion across centuries; example: Herbert Jevons predicted British coal use peaking around 2100 at 400 million tonnes per year, then stabilizing at 300 million tonnes until 2200. - The term originates in nuclear physics and was adopted by nuclear-power advocates (fast breeder reactors) envisioning a centuries-long evolution from finite to almost infinite energy. - The 1970s energy crisis popularized “energy transition” as a flexible slogan accommodating visions from nuclear fixes to coal revival, solar utopias, steady-state economies, and environmental critiques. - When global warming emerged as a political issue in the late 1970s, frameworks developed for the energy crisis (e.g., carbon pricing) were retooled to address it, despite a mismatch: scarcity-driven transitions differ fundamentally from climate-driven required rapid phaseout of fossil fuels. ### Present assessment: lifecycle emissions and the pillars of industrial civilization - The clear-cut dichotomy between “green” and “brown” technologies is contradicted by lifecycle emissions assessment and embedded materials. - It is implausible that renewables could competitively produce, on a global scale and by 2050, the "pillars of industrial civilization" (steel, cement, plastics and fertilizers). - Example lifecycle and system burdens: - Refrigerants estimated at 3 percent of global emissions and 7 percent of global electricity consumption in 2024. - IEA: impacts of climate change contributed about half of the increase in global energy demand in 2024; space cooling is a leading driver. - IEA expects power demand to grow at 4 percent per year through 2027. - Material and intermediate-input linkages (e.g., electric vehicles requiring steel, plastic) challenge a strict green/brown sector dichotomy. ### Hypotheses about the future (three core hypotheses) - First: Technological sprawl is likely to persist. - Improvements in technical efficiency have overwhelmingly generated technological sprawl rather than conservation of materials and energy. - Example: efficiency gains in computing enabled a surge in per-capita number of smartphones, tablets and other smart appliances. - In 2025, rapid growth in artificial intelligence induced sharp growth in data centers, combined with electrification-driven power demand growth, led to strong growth in gas demand, a shortage in gas turbines, and a switch to coal for power generation in many countries. - Preliminary evidence: reduced sulfur dioxide emissions from ships may generate additional global warming of 0.2C per decade (Hansen et al. 2023); similar effects noted for reduced sulfate emissions from land transportation. - Second: Upward pressure on global emissions from carbon-intensive adaptation investments. - Sea walls and irrigation systems require large quantities of concrete. - Air conditioning requires considerable energy and generates large amounts of powerful GHGs (refrigerants: 3 percent of global emissions; 7 percent of global electricity consumption in 2024). - Desalination has large energy needs and generates toxic brine that adversely impacts marine habitats’ ability to act as a carbon sink. - With climate impacts contributing about half of the increase in global energy demand in 2024, positive feedback loops between climate change and emissions will strengthen, stymieing decarbonization. - Third: Political economy factors likely to continue constraining mitigation. - Slow decarbonization, partly due to the Generalized Jevons Paradox (GJP), reinforces channels that entrench fossil fuel production and consumption via powerful vested interests (large corporations including ICE-centered automobile companies, fertilizer producers, state-owned and international fossil fuel companies). ### Empirical historical measures of efficiency and output - Between 1900 and 1950 in the U.S.: - The power of all available machines was multiplied by 70. - CO2 emissions were multiplied by only four. - Rapid efficiency gains between 1915 and 1930: - Rail freight reduced emissions per ton-kilometer by 30 percent. - Blast furnaces reduced the carbon intensity per ton of steel by 22 percent. - Cement plants reduced the carbon intensity per ton of cement by 27 percent. - Power plants divided carbon intensity per kWh by five. ### Theory: limits of the Energy Transition (ET) framework - Key implicit ET assumptions identified: - Energy and material dynamics are governed by substitution through diffusion of innovations. - Technological dynamics drive changes in consumption of materials. - A clear-cut dichotomy exists between “green” and “brown” technologies or “low-carbon” and “high-carbon” sectors. - Evidence contradicting these assumptions: - Global energy and material dynamics exhibit growing embeddedness of emissions and materials rather than substitution-driven declines. - Different primary energies are intertwined and interdependent; the global economy features accumulation of primary energies and materials over time. - High-carbon industries remain involved as intermediate inputs across production networks, undermining a strict green/brown dichotomy. ### Key concepts introduced - Uncertainty paradigm - Energy and material dynamics encompass deep structural uncertainty (“unknown unknowns”) with “unlimited downside liability on planetary damages” (Weitzman 2011). - This justifies precautionary decarbonization and the need to understand economy–biophysical interactions. - Jevons Paradox - Historical observation: greater efficiency in coal use increased rather than decreased coal use. - Mechanisms: efficiency lowers energy prices, increasing profitability in energy-intensive industries (reallocation of capital) and lowering consumer prices (raising aggregate demand); both can increase energy consumption. - Empirical support: documented strong rebound effects (e.g., Rausch and Schwerin 2018; Sorrell et al. 2009; Magee and Devezas 2017; Brockway et al. 2021). - Energy Symbiosis - Describes network of material interdependencies among technologies: emissions and materials embedded in products/materials. - Example: insulation efficiency gains reduce heating energy per home (rebound), but drive demand for resins, glass wool, polyurethane foam and ultimately fossil fuels for their production. - Symbiosis operates upstream of investment decisions and captures persistent complementarities and path dependency in materials and inputs. - Generalized Jevons Paradox (GJP) - Macro-level, long-term synthesis of Jevons Paradox and energy symbiosis, adding institutional and political economy mechanisms. - Empirical pattern: lower global carbon and material intensity (dollars per tCO2e or kilogram) has historically led to higher global GHG emissions and greater material footprints, or at least decline too slow to meet environmental goals. - GJP emphasizes drivers of investment and institutional determinants shaping productive structure, path dependence, and lock-in. ### Political economy drivers of the GJP - Core actors hypothesized to drive institutional determinants of the GJP: - Multinational corporations (MNCs) - States - Large financial actors (“finance”) - Institutional factors influenced by these actors: 1. Relative expected sectoral profitability, rent seeking, and market power. 2. Geoeconomic pressures (political and geopolitical power). 3. International trade arrangements. 4. International financial architecture. - Role of MNCs - Objective: maintain high levels of profitability; expected profitability shapes investment decisions. - Low levels of investment in renewable energy in developed economies partly due to relatively low expected profitability in these sectors. - High investment in some countries (e.g., China) associated with significant subsidies forcing diffusion and cost reduction. - Investment costs and long-term return profiles reduce commercial attractiveness of many low-carbon technologies even when their costs are lower than fossil equivalents. - Asymmetry in sectoral profitability is negatively related to sectors’ potential role in decarbonization. - Some decarbonizing activities (renewable energy, shared/active mobility, frugal innovation) have inherently limited profit-generation potential because they are not linked to scarce, rent-generating resources or are not conducive to property-rights-driven profits. *Source: Authors. (Content excerpt from "2. Embedded Emissions and Materials in 1930.")* ### Box 1. Expected Profitability and Investment in the Solar Industry ### Box 1. Expected Profitability and Investment in the Solar Industry ### Solar industry structure and expected profitability - The solar industry has low barriers to entry and is thus highly competitive, making expected profitability relatively low and uncertain. - Business model characteristics that depress expected profitability: - Upfront sale of infrastructure (for developing solar plants). - Highly volatile spot electricity prices, with negative electricity prices being mainly associated with renewables. - The low expected profitability is reflected in policy actions: the low and uncertain profitability of the solar industry is reflected in the actions taken by the Chinese government in September 2025 to address a surge in losses in the industry (White et al. 2025). - Empirical observation cited: Christophers (2023) describes a “cannibalization effect” whereby renewable electricity generators “undermine the conditions of their own future viability and growth precisely by virtue of their own success in displacing more expensive, conventional generating technologies and thereby reducing the price of electricity.” ### Comparison with oil and gas and sensitivity to cost of capital - Oil and gas industry features: - High barriers to entry and oligopolistic market structure. - Business model involves continued sales of scarce raw materials. - Highly profitable (Christophers 2023, p. 218). - Low-carbon technologies have a much higher ratio of capital expenditure to operational expenditure than fossil fuels, and hence much greater sensitivity to the cost of capital. - The oil and gas industry’s accumulation of very large profits in recent years (partly because of the energy crisis of 2022-2023) means it relies relatively little on external financing. ### Profitability imperative and investment decisions - The objective of maintaining capital accumulation at a minimum rate – firms’ profitability imperative – plays an important role in financing and investment decisions across countries and sectors. - Evidence indicates expected profitability plays a more important role than costs in determining investment in renewable energy, constraining its deployment (Christophers 2023, pp. 212-13). ### Planetary boundaries, profitable alternatives, and substitution - Out of nine planetary boundaries, only two have not been breached: increased atmospheric aerosol loading and stratospheric ozone depletion (Stockholm Resilience Centre 2025). - A distinctive feature of these two boundaries: profitable alternatives to the technologies driving the boundary existed (or emerged) and enabled large-scale substitution: - Aerosol loading: profitable alternative technologies emerged gradually over the last 40 years and are only now driving improvements. - Ozone depletion: substitution products were available in the 1980s for a small number of industries (air conditioning, refrigeration, propellants), limiting substitution costs. - At the aggregate level, these patterns are consistent with the hypothesis that expected profitability is a key determinant of the ability to limit global environmental impacts through large-scale technological substitution. ### Rent seeking, market concentration, and structural constraints - Rent seeking and market power orient profitability and investment; market concentration signals undue profits. - Supply concentration: the top three producing countries account for about 65 percent, 50 percent, and 70 percent of the global production of agriculture, energy, and minerals (IMF 2023, p. 74). - Near-universal availability of sunlight and kinetic wind energy creates a structural constraint on the profitability of renewable energy relative to concentrated supply sectors (coal, oil, gas, minerals). ### Domestic vested interests and EMDEs - In EMDEs, large importing companies often constitute powerful vested interests against decarbonizing the productive structure because imported goods (internal combustion engine vehicles, gasoline, agricultural products) have large and hard or impossible to abate carbon/emission footprints. ### Role of states and geoeconomic dynamics - Governments shape institutional drivers of the Generalized Jevons Paradox (GJP) and increasingly use geoeconomic pressures to affect profitability. - Geoeconomic definitions cited: - “an area of study of economics and political science that focuses on how governments use their country’s economic strength to exert influence on foreign entities” (Clayton et al. 2025). - Geoeconomic fragmentation: “a policy-driven reversal of global economic integration, guided by strategic considerations such as national security, sovereignty, autonomy, or economic rivalry” (Weber et al. 2025). - Example of state-influenced low-carbon FDI: since 2022, China’s low-carbon technology foreign direct investments (in batteries, solar, wind, new energy vehicles, green hydrogen) have amounted to over US$220 billion in 54 countries (Xue and Larsen 2025). ### Military spending, demand, and emissions - Military spending as profits source and demand driver: - Military spending more than doubled between 2000 and 2021, rising from around US$1 trillion to US$2.1 trillion per year. - It grew by close to 30 percent between 2021 and 2024, reaching $US2.7 trillion in 2024 (Liang et al. 2025). - The EU’s ReArm Europe program, launched in March 2025, implies an increase in military spending of around $US765 billion by 2029. - The defense industry would be the fourth largest global emitter if it were a country; military dynamism could increase the emission intensity of structural economic change and crowd out lower-emission spending. ### Legal instruments, private entrenchment, and trade power relations - Private corporations access state-backed legal instruments (“legal subsidies”) that entrench interests (Pistor 2025). Examples include: - Regional and bilateral investment treaties. - Legal modules: property rights and collateral, contract, trust, corporate and bankruptcy law. - Investor State Dispute Settlement (ISDS) exposes governments to legal action when implementing regulatory measures. - Trade power relations shape investment strategies and domestic political economy; multilateral predictability is fading and ad hoc bilateral arrangements proliferate. - Environmental considerations have been largely absent from trade agreements, except for the EU Carbon Border Adjustment Mechanism (Magacho et al. 2023). - Stylized facts on trade and environment (summary cited): polluting industries are more exposed to trade; negative environmental bias in trade policy; polluting industries upstream in value chains; traded products embed a fourth to a third of global emissions; advanced economies outsource greenhouse gas emissions (Copeland et al. 2022; Shapiro 2021). ### Finance, monetary policy, and expected profitability - Finance shapes investment via creation, trading and enforcement of financial claims; banks create the majority of money (McLeay et al. 2014); non-bank sector uses leverage extensively (BIS 2024). - Under bailout expectations, firms may engage in correlated behavior and large investments in carbon-intensive sectors (Stern and Stiglitz 2021). - Short-term profit expectations of financial institutions supported oil and gas investment and reduced renewables investment during the 2022 energy crisis (Christophers 2022). - Central bank and regulatory frameworks were largely designed before climate and fossil-fuel reduction were core mandates, though some central banks are beginning to incorporate climate considerations (ECB 2025; Schnabel 2020; NGFS 2024a; FSB 2025). - Central bank crisis-era asset purchases can shift profitability expectations by supporting firms’ debt and signaling support for credit-weakened firms; preliminary estimates (Espagne et al. 2023) suggest the implied global temperature rise of purchases made under crisis-era asset purchase programs of two major central banks during the COVID-19 crisis stood at 2.3°C to 3.1°C. - Financialization has raised profitability requirements and increased the cost of financial capital (Cordonnier et al. 2019; Lazonick and O’Sullivan 2000). - Finance affects corporate governance and therefore drives the GJP by shaping expected profitability in favor of carbon-intensive sectors. ### Policy role and potential to shift institutional drivers - Policies can play a central role in shaping or changing institutional drivers of the GJP; interdependencies can evolve over time beyond pure technological innovation. - Policy action can help determine the boundary between the ET (environmental transition) and the GJP as the framework of reference. ### Elements of formalization and conceptual implication - Preliminary formalization efforts aim to incorporate the Generalized Jevons Paradox into climate, energy and resource economics. - A key challenge to standard modeling assumptions is the high substitutability among inputs; empirical evidence suggests low substitutability between clean and dirty energies at the level of the world energy system. - Main implication: the GJP does not depend on the evolution of the cost of individual energy sources or technologies, or on the carbon intensity of the economy; declines in the cost of low-carbon technologies (notably solar and wind) can lead to greater use of high-carbon energies by lowering energy costs, supporting aggregate demand, and increasing technological flexibility. - Economically, the GJP is reflected in an elasticity of substitution between “clean” and “dirty” energies that is close to zero at the level of the world energy system. *Source: IMF Working Paper — Box 1. Expected Profitability and Investment in the Solar Industry* ### Annex V provides additional elements of formalization from the perspective of the Kaya identity. ### Annex V — Additional elements of formalization from the perspective of the Kaya identity ### Formal model and elasticity evidence - Aggregate production function (two final-good sectors, clean 푌c and dirty 푌d): - 푌t = (푌ct(휀−1)/휀 + 푌dt(휀−1)/휀)휀/(휀−1) - ε ∈ (0, + ∞) is the elasticity of substitution between the two sectors. - Generalized Jevons Paradox (GJP) implication: - The GJP implies the two sectors are complements → ε < 1. - This contrasts with Acemoglu et al. (2012) assumption that substitutes (ε > 1) are empirically relevant. - Empirical evidence on energy price elasticities (meta-analyses): - Short-term price elasticity of energy demand: −0.21 (ranging from −0.29 to −0.02). - Long-term price elasticity of energy demand: −0.61 (ranging from −0.77 to −0.19). - Meta-analyses for gasoline price elasticity: between −0.6 and −0.3, implying ε < 1. - Key technical caveats: - Aggregating primary energy quantities by “clean” vs “dirty” is technically problematic due to heterogeneity in transformation efficiencies and energy-accounting conventions. - Difficulty obtaining meaningful estimates of substitution elasticities between energies and other production factors. - Food-system and biofuel emissions are very hard to abate yet often ignored or not explicitly modeled in economic models. - Footnote note: - There is a global research community working on improving representation of agriculture-related emissions in economic models (reference in source). ### Conceptual contrast: Energy Transition (ET) Framework vs Generalized Jevons Paradox (GJP) - Technical differences: - ET Framework: Energy transition via substitution; clean vs dirty technology; risk paradigm; high elasticity of substitution (weak sustainability). - GJP: Energy symbiosis with additive dynamics; dirty vs semi-clean technology; uncertainty paradigm; complementarity (strong sustainability). - ET views transitions as long phases of dominance of a primary energy source (relative changes in shares). GJP emphasizes continuous absolute additions in global energy consumption and steadily rising consumption of virtually all materials. - Institutional differences: - ET: Net-zero goals via technology diffusion and carbon pricing; emission abatement via substitution and electrification; key mechanism is change in relative prices. - GJP: Emphasis on managing global/sectoral/group energy and material demand; role of institutions in investment choices and welfare trade-offs; primary management focus is global energy and material demand. - Ethical differences: - ET: Optimism about technological breakthroughs; gives significant role to breakthroughs in decarbonization strategy. - GJP: Precautionary principle emphasizing uncertainty and equity; ethically problematic to rely heavily on breakthroughs given deep uncertainty and catastrophic risk. - Scope and risk paradigms: - ET: Climate change focus; risk paradigm with optimization (e.g., IAMs, Social Cost of Carbon). - GJP: Planetary boundaries plus other environmental crises; deep, fat-tailed uncertainty with non-linear tipping points. ### Implications for economic research — priority areas - Overview: - Three priority research areas guided by Table 2: (1) detailed analysis of materials in supply chains and vulnerabilities; (2) political economy of material flows; (3) multidimensional welfare analysis of scenarios. - Policies can and should shape institutional drivers of the GJP; interdependencies can evolve for reasons beyond technological innovation. - 1) Detailed analysis of the role of materials in supply chain vulnerabilities - The GJP reveals a systemic nexus between materials, energies, and emissions leading to macroeconomic criticality of interdependencies. - Consequences of interdependencies: - Amplification of physical and economic/financial shocks. - Larger and more frequent shocks raise prices via supply shortages, damaged agricultural/industrial production, and trade disruptions. - Energy price shocks have a large effect on inflation through sectoral channels; energy dependence affects inflation passthrough. - Large physical shocks imply large supply constraints; price signals’ ability to enable adjustment would be limited. - Demand constraints linked to green budgeting and green planning could magnify effects. - Repeated or simultaneous shocks compound impacts; borrowing costs and debt levels could increase (due to lower growth prospects, weaker fiscal positions, bailout costs), challenging central banks’ price stability objective. - Research tools and policy uses: - Input-output analysis to map uses of materials (e.g., sulfur monoxide, oil, ammonia) and associated vulnerabilities. - Inform industrial and green industrial policies to target public support, reduce reliance on energy imports, and identify interdependencies and strategies to change productive structures. - Input-output analysis to envisage technologically-constrained pathways and assess sectoral resource distributions. - 2) Political economy of material flows - Need for political-economy analysis of coalitions supporting institutional arrangements that drive the GJP: rent-seeking, market power, geoeconomic pressure, international trade dynamics, and international financial architecture. - Research goals: - Understand how vested interests and sociopolitical groups form and organize and how accumulation of political power entrenches carbon-intensive material flows. - Analyze how economic and social change affects political power and the emergence or decline of sectors (demography, technical change, social contract impacts). - Articulate material flow analysis with political economy of institutional architecture, using concepts like “historical viability regimes” to study how institutions and social groups historically shaped energy and material flows. - Develop structural models that endogenize impacts of large-scale policy and regulatory changes. - 3) Multidimensional welfare analysis of scenarios - GJP implication: persistence (“stickiness”) of fossil fuels in the energy system and inertia in global emissions; rebound effects amplify stickiness. - Policy and distributional implications: - Hard-to-abate emissions (cement, steel, other heavy industries) are crucial for development (housing, roads, renewable energy); absence of affordable zero-emission options may require careful consumption distribution with priority for low-income countries, raising difficult distributional questions across and within countries. - Research needed on differential utility of fossil fuels across sectors; current models may not capture nuanced impacts of declining activity in hard-to-abate sectors. - Policymakers need tools to quantify sector-specific decarbonization effects and economy resilience to potential sector downsizing. - Incorporate non-monetary evaluations of CO₂ utility grounded in welfare economics to better reflect societal well-being. - Existing IAMs that explore significant energy demand shifts operate at highly aggregated levels, limiting usefulness for targeted policy design. - Develop models enabling sophisticated distributional analysis of decarbonization scenarios: - Move beyond Pareto optimality assumptions common in net-zero scenarios to analyze material and emissions transfers between and within countries and across income groups. - Use such models to evaluate fairness in distribution of decarbonization efforts and outcomes. *Source: IMF Working Paper — Annex V content as provided.* ### Conclusion ### Conclusion ### Main concept: The Generalized Jevons Paradox (GJP) - Definition: Higher energy efficiency (and higher efficiency in the use of other materials) at the level of the world economy leads to higher — rather than lower — total energy use (and total use of other materials). - Counterintuitive implication: Lower carbon intensity of the world economy (emissions per unit of GDP) is likely to lead, in a central scenario, to higher (or, at a minimum, persistently high) global emissions. ### Relationship to existing concepts - Jevons Paradox: Saving energy tends to lead to more energy consumption; GJP augments this by emphasizing systemic, structural, and institutional mechanisms that generate rebound effects. - Embedded emissions and materials (energy symbiosis): Network of material interdependencies among technologies; path dependency and technological complementarities in the long run. - Distinction: The Jevons Paradox focuses on micro and macro effects from efficiency gains; embeddedness focuses on technological complementarities; the GJP focuses on root causes in institutional drivers of investment and captures dynamic interactions among technologies, energy, materials, and institutions. ### Stylized facts emphasized by the GJP - World energy system governed by additive dynamics and energy symbiosis: different forms of energy accumulate rather than fully substitute one another. - Technological and material dynamics can differ significantly: rapid deployment of low-carbon technologies is compatible with continued (or increased) fossil fuel consumption. - Historical pattern: accumulation, in absolute terms, of different energies and materials. ### Implications for the Energy Transition (ET) framework - Two core ET assumptions challenged: 1. Substitution dynamics (technologies substituting other energy forms via diffusion). 2. Broad equivalence of technological change and material change (deployment of low-carbon technologies implies commensurate decrease in fossil fuel consumption). - Evidence suggests: - Technological and material dynamics are not equivalent. - Elasticity of substitution between “clean” and “dirty” technologies is relatively low (and in some cases zero). - In absence of binding constraints on energy consumption, there is a macro-scale rebound effect associated with increased energy efficiency. - Consequence: Global productive structure tends to become larger and more complex over time, entrenching fossil fuels and producing “stickiness” in global emissions. ### Policy-relevant cautions and prospects - Need for caution when applying standard ET framework; importance of reducing energy demand to achieve significant and sustained decarbonization over the next decades. - Possibility of sustained decline in global emissions driven by a major economy is not ruled out but: - Example references: Myllyvirta 2025 and Xue and Larsen 2025 on China’s emissions profile since 2024 and its very high FDI in low-carbon sectors since 2022. - Such a hypothesis rests on strong assumptions: China’s carbon intensity remains very high, and any global dynamic would probably unfold over several decades. - Recommend adopting an uncertainty paradigm rather than assuming deterministic substitution-to-decarbonization pathways. ### Political economy drivers and key actors - Four drivers shaping investment choices and thus the GJP: 1. Rent seeking and market power. 2. Geoeconomic pressures. 3. International trade arrangements. 4. Financial flows. - Three sets of actors that shape these drivers: - Large companies and multinational corporations. - States. - Large financial actors. - Likely impacts of GJP-informed dynamics: - Perpetuation of large energy rents. - Persistence of “technological sprawl” — continued diffusion of a wide range of technologies in consumption patterns. - Large emissions tied to construction, operation and maintenance of infrastructure (including for climate adaptation and mitigation). - Continued political economy constraints to climate change mitigation. ### Empirical evidence summarized (selected quantitative findings) - Regression analyses show statistically significant positive relationships between resource-use efficiency and total resource use across multiple domains, consistent with the GJP. Estimates are similar across OLS, median quantile regression, and Theil-Sen estimator. - Energy efficiency and total emissions (1900-2023): - OLS coefficient on energy efficiency: 16.94*** (standard error 0.766), Constant -3.572* (1.889), N 68, R2 0.881. - Quantile regression coefficient: 16.13*** (0.887). - Theil-Sen slope: 17.265 (95% CI Lower 15.808, Upper 19.168); intercept -1.902. - Water use efficiency and total water use (1960-2014): - OLS coefficient: 198.6*** (12.42), Constant 1,006*** (134.9), N 52, R2 0.836. - Quantile regression coefficient: 193.9*** (18.17). - Theil-Sen slope: 221.015 (95% CI Lower 187.555, Upper 259.193); intercept 1046.28. - Agriculture nitrogen use efficiency and total nitrogen used (sample restricted to 1985-2023): - OLS coefficient: 7,473*** (536.2), Constant -76,229 (49,700), N 39, R2 0.84. - Quantile regression coefficient: 7,175*** (844.6); Constant -50,204 (78,288). - Theil-Sen slope: 7420.703 (95% CI Lower 6164.207, Upper 8752.042); intercept -54185.703. - Electricity share in final energy consumption and total energy/fossil fuel consumption (1971-2023): - Effect on total energy consumption (EJ): - OLS coefficient: 2,123*** (50.95), Constant -23.30*** (7.816), N 53, R2 0.971. - Quantile regression: 2,092*** (92.35), Constant -17.39 (14.17). - Theil-Sen slope: 2129.455 (95% CI Lower 1993.782, Upper 2238.717); intercept -39.242. - Effect on fossil fuel consumption (EJ): - OLS coefficient: 1,148*** (39.38), Constant 23.60*** (6.042), N 53, R2 0.943. - Quantile regression: 1,135*** (71.40), Constant 26.88** (10.95). - Theil-Sen slope: 1158.363 (95% CI Lower 1055.637, Upper 1241.488); intercept 11.418. ### International trade and embedded footprints - Sankey-diagram evidence (across six planetary boundaries) shows: - High- and middle-income countries are main drivers of import consumption pressure; flows originate largely in middle- and low-income regions (notably Latin America and the Caribbean, and East Asia and the Pacific) and terminate in high-income regions (Europe, North America, high-income Asian countries). - Many lower-income regions export significantly more ecological pressure than they import; Latin America exports over four times more pressures than it imports for the biosphere integrity boundary. - Result: wealthy countries can offload environmental degradation elsewhere, contributing to the stickiness of the GJP and masking the true footprint of consumption. ### Research agenda — three main avenues - Detailed analysis of the role of various materials in creating supply-chain vulnerabilities, given deep entwinement between energies and materials and intensifying physical shocks. - Political economy of material flows: identify political coalitions and institutional arrangements that shape GJP drivers and orientations of material flows. - Multidimensional welfare analysis of scenarios: study the differential usefulness of fossil fuels across economic sectors given likely persistence of fossil fuels in the global energy system. *Source: IMF Working Paper — Conclusion section.* ### References ### References ### Directed technical change and innovation - Acemoglu, Daron, Philippe Aghion, Leonardo Bursztyn, and David Hemous. 2012. “The Environment and Directed Technical Change.” American Economic Review 102 (1): 131–66. https://doi.org/10.1257/aer.102.1.131. - Acemoglu, Daron, Philippe Aghion, Lint Barrage, and David Hémous. 2023. “Green innovation and the transition toward a clean economy.” No. WP23-14. Washington, DC: Peterson Institute for International Economics. - Aghion, Philippe, Céline Antonin, and Simon Bunel. 2021. The Power of Creative Destruction: Economic Upheaval and the Wealth of Nations. Cambridge, MA: Belknap Press. - Aghion, Philippe, Roland Bénabou, Ralf Martin, and Alexandra Roulet. 2023. "Environmental preferences and technological choices: Is market competition clean or dirty?" American Economic Review: Insights 5(1): 1-19. - Aghion, Philippe, Timo Boppart, Michael Peters, Matthew Schwartzman, and Fabrizio Zilibotti. 2025. “A theory of endogenous degrowth and environmental sustainability.” NBER Working Paper No. 33634. Cambridge, MA: National Bureau of Economic Research. https://doi.org/10.3386/w33634. ### Macroeconomics, climate models, and social cost of carbon - Barrage, Lint, and William Nordhaus. 2024. “Policies, Projections, and the Social Cost of Carbon: Results from the DICE-2023 Model.” Proceedings of the National Academy of Sciences 121 (13): e2312030121. https://doi.org/10.1073/pnas.2312030121. - Bilal, Adrien, and James H Stock. 2025. Macroeconomics and Climate Change. Mimeo. - Black, Simon, Dora Benedek, Ian Parry, Nate Veron-Lin, and Sunalika Singh. 2025. “Cutting Emissions, Securing Energy: A Macroeconomic Assessment for COP30.” IMF Working Paper No. 2025/245. Washington, DC: International Monetary Fund. ### Energy history, markets, and geopolitics - Black, Brian C. 2022. To Have and Have Not: Energy in World History. London: Rowman & Littlefield. - Black, Brian C. 2023. Energy Revolutions: A History. Hoboken, NJ: Wiley. - Alvarez, Jorge, Mehdi Benatiya Andaloussi, Chiara Maggi, Alexandre Sollaci, Martin Stuermer, and Petia Topalova. 2025. “Geoeconomic Fragmentation and Commodity Markets.” IMF Working Paper No. 2023/201. Washington, DC: International Monetary Fund. - Alvarez, Jorge, and Thomas Kroen. 2025. "The Energy Origins of the Global Inflation Surge." IMF Working Paper No. 2025/144. Washington, DC: International Monetary Fund. - Bachmann, Ruediger, David Baqaee, Christian Bayer, Moritz Kuhn, Andreas Löschel, Benjamin Moll, Andreas Peichl, Karen Pittel, and Moritz Schularick. 2024. "What if? The macroeconomic and distributional effects for Germany of a stop of energy imports from Russia." Economica 91 (364): 1157-1200. - Bogmans, Christian, Ervin Prifti, Ivan Petrella, Martin Stuermer, and Andrea Pescatori. 2024. “The Power of Prices: How Fast Do Commodity Markets Adjust to Shocks?” IMF Working Paper No. 2024/077. Washington, DC: International Monetary Fund. - Associated Press. 2018. “Global Electric Motor Market Analysis, Trends, Share, Size, Growth, Opportunity & Forecast 2018-2023, With an Expected CAGR of 10%.” ### Environmental economics, production, and externalities - Ayres, Robert U., and Allen V. Kneese. 1969. “Production, Consumption, and Externalities.” The American Economic Review 59 (3): 282–97. ### Policy, institutions, and financial sector perspectives - BIS. 2024. Annual Economic Report 2024. Basel: Bank for International Settlements. - Blinder, Alan S. 2022. A Monetary and Fiscal History of the United States, 1961-2021. Princeton, NJ: Princeton University Press. - Bolton, Patrick, Morgan Després, Pereira da Silva, Luiz Awazu, Frédéric Samama, and Romain Svartzman. (entry listed in source; full citation text continues beyond provided excerpt) ### Other books and works in the list - Aglietta, Michel, and Etienne Espagne. 2024. Pour une écologie politique: Au-delà du capitalocène. Paris : Odile Jacob. - Amable, Bruno, and Stefano Palombiarini. 2024. Bloc sociaux et domination. Paris: Raisons d’agir. *References — wpiea2026157-source-pdf* ### 2020. The Green Swan. Central Banking and Financial Stability in the Age of Climate Change. Basel and ### The Generalized Jevons Paradox and the Future of Energy ### Overview and scope reflected in references - The bibliography compiles multidisciplinary literature spanning economics, energy history, environmental science, climate policy, finance, and central banking relevant to the generalized Jevons paradox and energy futures. - Key institutional sources cited include Bank for International Settlements, Banque de France, International Monetary Fund, International Energy Agency (IEA), Network of Central Banks and Financial Supervisors for Greening the Financial System (NGFS), Financial Stability Board (FSB), Intergovernmental Panel on Climate Change (IPCC), and other policy and academic outlets. ### Core thematic areas represented - Energy efficiency, rebound effects, and the Jevons paradox - Foundational and recent works: Jevons, William Stanley. 1865. The Coal Question; Khazzoom, J. Daniel. 1980; Gillingham, Kenneth, David Rapson, and Gernot Wagner. 2016; Sorrell, Steve, John Dimitropoulos, and Matt Sommerville. 2009; Fix, Blair. 2024. “A Tour of the Jevons Paradox: How Energy Efficiency Backfires.” - Empirical and theoretical treatments of rebound and economy-wide effects: Brockway et al. 2021; Heun, Semieniuk, and Brockway. 2025; Rausch and Schwerin. 2018; Magee and Devezas. 2017. - Energy transitions, technological change, and historical perspectives - Historical and technological analyses: Smil (2010, 2017, 2022, 2024), Grubler 2012, Wrigley 2011, Penna 2021, Marchetti 1977. - Modeling and scenarios: Grubler et al.; Lefèvre 2024; Semieniuk et al. 2021; Lilliestam, Patt, and Bersalli 2020. - Climate risks, carbon budgets, and biosphere feedbacks - Climate science and carbon sink dynamics: IPCC 2021; IPCC 2022; Pan et al. 2024; Ke et al. 2024; Carle et al. 2025; Curran and Curran 2025. - Planetary boundaries and biodiversity linkages: Richardson et al. 2023; IPBES 2024; Stockholm Resilience Centre 2025. - Macroeconomic, trade, and industrial policy implications - Macroeconomic modeling and policy tools: Fornaro, Guerrieri, and Reichlin. 2025; Mitra et al. 2025; Miranda-Pinto et al. 2024; Magacho et al. 2023; IMF World Economic Outlook chapters (2023, 2024a, 2024b, 2025a, 2025c) and Global Financial Stability Report 2025b. - Industrial and trade policy: Rodrik 2014, 2018; Juhász, Lane, and Rodrik 2024; Xue and Larsen 2025; Tubiana et al. 2025. - Financial stability, central banking, and geoeconomic risks of transition - Central banking and supervisory responses: The Green Swan. 2020 (BIS and Banque de France); Schnabel 2020; NGFS 2023, 2024a, 2024b; FSB 2025; ECB 2025; Oman, Salin, and Svartzman 2024; Weber et al. 2025. - Cross-border and geoeconomic fragmentation risks: Espagne et al. 2023; Weber et al. 2025; Clayton, Maggiori, and Schreger 2025; Braun and Koddenbrock 2023. - Resource constraints, material flows, and critical metals - Resource economics and criticality analyses: Hotelling 1931; Nordhaus (1974, 1992, 1994); Bucciarelli, Hache, and Mignon 2025; Crippa et al. 2021; Yokoi, Watari, and Motoshita 2022; Ren et al. 2022. ### Representative influential citations (selected with years retained) - The Green Swan. 2020. Bank for International Settlements and Banque de France. - Jevons, William Stanley. 1865. The Coal Question. - IPCC. 2021. Summary for Policymakers. Climate Change 2021: The Physical Science Basis. - IPCC. 2022. Climate Change 2022: Mitigation of Climate Change. - Gillingham, Kenneth, David Rapson, and Gernot Wagner. 2016. “The Rebound Effect and Energy Efficiency Policy.” Review of Environmental Economics and Policy 10 (1): 68–88. - Brockway, Paul E., Steve Sorrell, Gregor Semieniuk, Matthew Kuperus Heun, and Victor Court. 2021. “Energy Efficiency and Economy-Wide Rebound Effects: A Review of the Evidence and Its Implications.” Renewable and Sustainable Energy Reviews 141 (May): 110781. - NGFS. 2023. “Recommendations toward the Development of Scenarios for Assessing Nature-Related Economic and Financial Risks.” - IMF. 2025a. World Economic Outlook, Chapter 1, “Global Prospects and Policies.” Washington, DC: International Monetary Fund. - IEA. 2024. World Energy Outlook 2024. ### Policy and analytical strands suggested by the references - Integrate rebound-effect evidence into energy-efficiency policy design and climate mitigation pathways (references include Gillingham et al. 2016; Brockway et al. 2021; Fix 2024). - Incorporate climate-science updates on carbon sinks, committed emissions, and planetary boundaries into macroeconomic and financial-stability models (references include IPCC 2021; IPCC 2022; Tong et al. 2019; Pan et al. 2024; Ke et al. 2024). - Central banks and supervisors to adapt operations and risk assessment frameworks to climate-related transition and physical risks (references include The Green Swan 2020; NGFS 2023; NGFS 2024a; NGFS 2024b; FSB 2025; ECB 2025). - Account for material and critical-metal constraints in planning decarbonization and electrification strategies (references include Bucciarelli et al. 2025; Yokoi et al. 2022; IEA 2015). - Consider geoeconomic fragmentation and trade impacts of the green transition in policy design (references include Espagne et al. 2023; Clayton, Maggiori, and Schreger 2025; Xue and Larsen 2025). *Working Paper No. WP/2026/157* --- _Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026157-source-pdf.pdf_