## rastea ## Source details **Canonical URL:** [rastea](https://www.imf.org/-/media/files/publications/dp/2026/english/rastea.pdf) ## Other formats - [Markdown version](/-/media/files/publications/dp/2026/english/rastea.pdf.md) - [Structured JSON version](/-/media/files/publications/dp/2026/english/rastea.pdf.json) --- ### Executive Summary — Overview and context - The period labeled the “Great Moderation” (mid-1980s to 2007) was interrupted by the global financial crisis and followed by a “Great Turbulence” with structural shifts, frequent geopolitical and supply shocks, and volatile trade and capital flows. - Four accelerating structural transformations examined: the demographic transition, climate change, geoeconomic fragmentation, and artificial intelligence (AI). - Many countries entered this period with historically high public debt levels, in part because of the COVID-19 pandemic, reducing capacity to support stability and adaptation. - Central question: are current macro policy frameworks fit for purpose in this “new normal”? ### Estimated growth impacts of structural transformations (moderate scenario, 2026–35) - Global context: - Comparison baseline: global average per capita growth (not including acute climate effects) of roughly 2 percent in 2023–24. - Sample coverage for growth analysis: 143 economies (34 AEs, 68 EMs, 41 LICs). - Advanced economies (AEs), average annual per capita GDP growth effects: - Demographics: −0.5 percentage points - Climate change (including chronic and acute effects): −0.3 percentage points - Geoeconomic fragmentation: −0.2 percentage points - Artificial intelligence: +0.3 percentage points - Estimated total impact (including both chronic and acute climate effects): −0.7 percentage points - Context: average annual per capita growth during 2015–24 = 1.6 percent - Emerging markets (EMs), average annual per capita GDP growth effects: - Geoeconomic fragmentation: −0.2 percentage points - Climate change (including chronic and acute effects): −0.4 percentage points - Demographics: −0.1 percentage points - Artificial intelligence: +0.2 percentage points - Estimated total impact (including both chronic and acute climate effects): −0.5 percentage points - Context: average annual per capita growth during 2015–24 = 2.4 percent - Low‑income countries (LICs), average annual per capita GDP growth effects: - Geoeconomic fragmentation: −0.2 percentage points - Demographics: +0.4 percentage points (conditional on productive employment and avoidance of capital shallowing) - AI: modest gains in medium term; median ~0.1 percentage points per year over 10 years - Climate: acute effects likely significant but hard to quantify for many LICs - Context: average annual per capita growth during 2015–24 = 2.3 percent - Interactions among megatrends: - Indicative estimate: selected interaction channels (higher barriers to trade and migration, slower AI diffusion) could shave an additional −0.4 percentage points of annual per capita growth on average in the moderate scenario. - Stylized aggregate (moderate scenario plus interactions, 2026–35): Moderate: —0.05; Interactions: —0.41; Total: —0.46 (average yearly impact on GDP per capita growth). ### Natural rates, term premia, and fiscal implications - Natural rates: - Likely to remain low in advanced economies. - Likely to decline further in emerging markets. - AI could raise natural rates under strong adoption (example: Melina and Villa (2025) scenario raises natural rates by 0.7 percentage points under a strong AI-led investment surge). - Term premia: - Signs of a potential reversal; term premia have risen somewhat alongside yield-curve steepening since early 2025. - Normalization of term premia to pre–global financial crisis levels could offset benefits from low natural rates, raising borrowing costs. - Estimated transformation-related fiscal spending needs: - Could average 3–4 percent of GDP annually over the next decade (2025–35) for combined demands (defense, climate adaptation, health and pensions, infrastructure, and human capital in source accounting). - Maximum sustainable debt (MSD) and fiscal space: - MSD levels likely to be lower because of weaker growth prospects, higher risk premia, and persistent spending pressures. - For many EMs and LICs, closing part of the tax gap can preserve fiscal space: - Closing up to half of the tax gap in EMs corresponds to median revenue gains of 2.6 percent of GDP annually. - Closing one-quarter of the tax gap in LICs corresponds to median revenue gains of 2 percent of GDP annually. - Such revenue mobilization could preserve fiscal space of approximately 23 percent of GDP over the next decade, on average (source context). ### Macro policy frameworks — pressures and operational needs - Fiscal policy frameworks — key pressures: - Continued elevated debt and higher borrowing costs. - Long-term spending pressures and more frequent supply shocks. - Structural shifts in the tax base and higher uncertainty about shock transmission. - Fiscal quantified need: - Transformation-related spending needs could average 3–4 percent of GDP annually over the next decade. - Monetary policy frameworks (inflation-targeting contexts emphasized) — key pressures: - More shock-prone environment sharpens inflation–output trade-offs. - Higher uncertainty about shock nature and transmission. - Risks of fiscal dominance and the effective lower bound (ELB). - Monetary operational needs: - Inflation‑targeting central banks may need to reassess when to “look through” supply shocks and make systematic use of alternative scenarios. - Use of some unconventional monetary policy (UMP) tools will require clearly defined escape clauses. - Preserve/strengthen central bank independence to ward off fiscal dominance. - External sector frameworks — key pressures: - Higher volatility of trade and capital flows. - Deepening financial frictions driven by geoeconomic fragmentation and climate shocks. - Possibly lower effectiveness of CFMs and a more fragmented global financial safety net (GFSN). - External operational needs: - Monetary and fiscal policies may need to be complemented more often by FX intervention (FXI) and macroprudential policy measures (MPMs). - Clear principles for use of FXI, reserve adequacy, and institutional capacity maintenance. ### Policy interactions and trade-offs (model-based findings) - Structural transformations exacerbate trade-offs between monetary, fiscal, and external policies: - Constrained fiscal space can weaken monetary transmission and amplify inflationary pressures. - Tighter monetary policy raises debt servicing costs, limiting fiscal options. - External constraints and financial stability risks can force simultaneous tightening even in adverse conditions. - Scenarios and model insights (QIPF and IPF variants): - Repeated adverse supply shocks increase inflation persistence and require larger policy-rate responses; example: secondary policy-rate response roughly 4 percentage points versus just over 2 in an illustrative repeated-shock case. - Fiscal consolidation can reduce the need for extreme monetary tightening and lower private consumption losses in some stylized AEs. - In vulnerable economies with high foreign debt and shallow FX markets, fiscal consolidation is especially effective at curbing inflationary pressures. - When fiscal space is limited and frameworks weak, large cost-push shocks produce worse inflation and debt outcomes even with aggressive tightening. ### Suggested adjustments to policy frameworks (high-level and specific) - Broad principles: - Balance credibility and flexibility; be more risk based and better integrated. - Strengthen ex ante robustness: stronger buffers, robust prudential safeguards, and credible institutions. - Emphasize policy interactions and coherent policy mixes while keeping distinct mandates. - Adopt state-contingent and risk-based policy design: routinely use risk scenarios alongside baseline forecasts and improve real-time shock identification. - Recognize limits of macro policies alone; complement with structural policies to boost diversification and adjustment mechanisms. - Fiscal policy possible adjustments: - Prioritize containing debt vulnerabilities; adopt risk-based medium-term debt anchors with safety margins. - Recalibrate fiscal rules to account for structural transformations; regular stress testing and error-correction mechanisms. - Reform entitlement programs where unsustainable; reduce spending rigidity and improve spending efficiency. - Redesign tax systems and upgrade social safety nets; consider rebalancing tax away from labor toward capital and corporate income where appropriate. - For EMs and LICs: pursue sustained revenue mobilization (examples: median revenue gains of 2.6 percent for EMs and 2 percent for LICs in modelled tax-gap closures). - Monetary policy (inflation-targeting) possible adjustments: - React more decisively to large supply shocks and establish conditions when “looking through” remains appropriate. - Make greater use of alternative scenarios and risk-based communications. - Preserve/strengthen central bank independence; pay greater attention to fiscal–monetary interactions. - Review ELB toolkit: consider more use of negative interest rates (e.g., Switzerland reached −0.75 percent historically), forward guidance with escape clauses, cautious use of QE with clearly defined exit clauses, tolerate small inflation undershooting when long-term expectations remain anchored. - External sector possible adjustments: - More systematic application of the Integrated Policy Framework (IPF). - Improve real-time shock identification and develop modalities for FXI, MPM, and CFMs (magnitude and instruments). - Prioritize developing domestic financial markets and adapting reserve adequacy metrics. - Reserve adequacy guidance: reserves in the range of 100–150 percent of the ARA metric considered adequate; ARA Metric = 0.05 × Exports + 0.05 × Broad Money + 0.3 × Short-Term Debt + 0.15 × Other Liabilities. ### Cross-cutting themes and institutional priorities - Reduced policy space and more frequent supply shocks increase the premium on ex ante robustness of policy frameworks. - Persistently high uncertainty motivates risk-based frameworks and state-contingent policy design; scenario analysis should become routine. - Structural policies are necessary to bolster resilience; without them, macro policies will face an ever-narrowing set of options to preserve stability. - Institutional priorities: - Strengthen central bank independence (operational, institutional, financial). - Strengthen fiscal institutions, fiscal councils, and public debt-management capacity. - Upgrade data and tools for real-time shock identification, scenario design, and reserve adequacy assessment. ### Purpose and next steps - The paper aims to stimulate discussion on possible adjustments to macro policy frameworks to make them more robust amid the Great Turbulence. - Provides guiding lines for more granular, country-level assessment of framework adjustments based on exposures and institutional characteristics. *Source: IMF DEPARTMENTAL PAPERS • Resilience amid Structural Transformations — Executive Summary and selected chapters as provided in the source content.* ### Executive Summary ................................................................................................1 ### Executive Summary ### Overview and context - The period labeled the “Great Moderation” (mid-1980s to 2007) was interrupted by the global financial crisis, followed by a “Great Turbulence” characterized by structural shifts, frequent geopolitical and supply shocks, and volatile trade and capital flows. - Four accelerating structural transformations examined: the demographic transition, climate change, geoeconomic fragmentation, and artificial intelligence (AI). - Many countries entered this period with historically high public debt levels, in part because of the COVID-19 pandemic, reducing capacity to support stability and adaptation. - Central question: are current macro policy frameworks fit for purpose in this “new normal”? ### Estimated growth impacts of structural transformations - Under a moderate-impact scenario, the combined effects of the transformations imply: - Advanced economies: average annual per capita GDP growth could decline by about 0.4 percentage points (not including acute climate effects) over the next decade; 0.7 with acute effects. - Emerging markets: average decline of 0.1 percentage points (not including acute climate effects); 0.5 with acute effects. - Low-income countries: possible average gain of 0.3 percentage points if the demographic dividend is effectively exploited. - Comparison baseline: global average per capita growth (not including acute climate effects) of roughly 2 percent in 2023–24. - The net effect could be much larger due to negative interactions between megatrends and more frequent/persistent supply shocks. - Natural rates: - Likely to remain low in advanced economies. - Likely to decline further in emerging markets. - Heightened uncertainty and rising public debt could put upward pressure on term premia, partially offsetting benefits of low neutral rates. ### Macro policy implications and constraints - Limited policy space and acute trade-offs make macro stabilization more challenging across three core frameworks: Fiscal policy frameworks - Key pressures: - Continued elevated debt and higher borrowing costs. - Long-term spending pressures. - More frequent supply shocks. - Higher uncertainty about the nature and transmission of shocks. - Structural shifts in the tax base. - Quantified need: - Transformation-related spending needs could average 3–4 percent of GDP annually over the next decade. - Fiscal risks: - Maximum sustainable debt (MSD) levels are likely to be lower because of weaker growth prospects, higher risk premia, and persistent spending pressures. - For countries near their tax capacity, fiscal space will remain highly constrained unless spending is reprioritized or entitlement programs are reformed. - For others, sustained revenue mobilization will be critical to preserve—or expand—fiscal space. Monetary policy frameworks (inflation-targeting contexts emphasized) - Key pressures: - A more shock-prone environment sharpens inflation–output trade-offs and reduces effectiveness of relying on a single baseline forecast. - Higher uncertainty about shock nature and transmission. - Risks of fiscal dominance and effective lower bound (ELB). - Operational needs: - Inflation-targeting central banks may need to reassess when to “look through” supply shocks and make more systematic use of alternative scenarios. - Use of some unconventional monetary policy (UMP) tools will require clearly defined escape clauses. - Preserving/strengthening central bank independence is essential to ward off fiscal dominance pressures. - Interactions between fiscal and monetary policy become more important. External sector policy frameworks - Key pressures: - Higher volatility of trade and capital flows. - Higher uncertainty about the nature and transmission of shocks. - Deepening financial frictions driven by geoeconomic fragmentation and climate shocks. - Possibly lower effectiveness of capital flow management measures (CFMs) and a more fragmented global financial safety net (GFSN). - Operational needs: - Monetary and fiscal policies may need to be complemented with additional tools—such as foreign exchange intervention (FXI) and macroprudential policy measures (MPMs)—more often. - Clear principles are needed to guide use of these tools, reserve adequacy assessments, and institutional capacity maintenance. ### Policy interactions and trade-offs - Structural transformations exacerbate trade-offs between monetary, fiscal, and external policies: - Constrained fiscal space can weaken monetary transmission and amplify inflationary pressures. - Tighter monetary policy raises debt servicing costs, limiting fiscal support options. - External constraints and financial stability risks can force simultaneous tightening even in adverse conditions. - Model-based simulations indicate careful coordination of policy levers can improve growth and inflation outcomes. - Example caution: fiscal stimulus to support an economy facing adverse supply shocks must be calibrated to avoid adding inflationary pressure that would require offsetting tighter monetary policy. ### Suggested adjustments to policy frameworks (high-level) - Broad principles: - Balance credibility and flexibility; be more risk based and better integrated. - Strengthen ex ante robustness: stronger buffers, robust prudential safeguards, and credible institutions. - Increase emphasis on policy interactions and coherent policy mixes while keeping distinct mandates. - Adopt state-contingent and risk-based policy design: routinely use risk scenarios alongside baseline forecasts and improve real-time shock identification. - Recognize limits of macro policies alone; complement with structural policies to boost diversification and adjustment mechanisms. - Fiscal policy possible adjustments: - Prioritize containing debt vulnerabilities. - Recalibrate fiscal rules to account for ongoing structural transformations. - Reform entitlement programs where unsustainable. - Recalibrate targeted fiscal support toolkits and automatic stabilizers. - Reduce spending rigidity. - Pay greater attention to effective fiscal and monetary interactions. - Redesign tax systems and upgrade social safety nets. - Monetary policy (inflation-targeting) possible adjustments: - React more decisively to large supply shocks and establish conditions when “looking through” remains appropriate. - Analyze how megatrends affect the inflation-output trade-off. - Make greater use of risk scenarios to formulate and communicate uncertainty and the central bank reaction function. - Preserve/strengthen central bank independence. - Pay greater attention to effective fiscal and monetary policy interactions. - Consider more use of negative interest rates, forward guidance with escape clauses, cautious use of QE, tolerate small inflation undershooting. - External sector possible adjustments: - More systematic application of the Integrated Policy Framework (IPF). - Improve real-time shock identification. - Develop effective modalities of FXI, MPM, and CFMs (magnitude and instruments). - Prioritize developing domestic financial markets, including adaptation to ongoing structural transformations. - Establish conditions for using FXI under more frequent shocks and higher uncertainty, and upgrade reserve adequacy metrics. ### Cross-cutting themes - Reduced policy space and more frequent supply shocks increase the premium on ex ante robustness of policy frameworks. - More limited scope for countercyclical policies and acute trade-offs require focus on policy interactions and coherent mixes; limited policy space may make deeper and longer recessions unavoidable in some episodes. - Persistently high uncertainty makes a stronger case for risk-based frameworks and state-contingent policy design. - Structural policies are necessary to bolster resilience; without them, macro policies will face an ever-narrowing set of options to preserve stability. ### Purpose and next steps - The paper aims to stimulate discussion on possible adjustments to macro policy frameworks to make them more robust amid the Great Turbulence. - It provides guiding lines for more granular, country-level assessment of framework adjustments based on exposures and institutional characteristics. *Source: IMF DEPARTMENTAL PAPERS • Resilience amid Structural Transformations — Executive Summary.* ### 1. Introduction ### 1. Introduction ### A. Structural Transformations — Key Channels and Findings - The global economy is being reshaped by four major structural transformations: demographics, climate change, geoeconomic fragmentation, and artificial intelligence (AI). - These transformations and their interactions are expected to: - Alter growth dynamics. - Amplify uncertainty. - Change the nature of shocks, leading to more frequent and persistent supply shocks and higher growth and inflation volatility. - Demographics: - Fertility rates have declined globally, dropping below the replacement level in many countries, while longevity has increased. - Global population is projected to increase until the 2080s, with significant variation across countries. - Example projections in the source: Korea’s population is projected to decrease by almost 60 percent by 2100; the population of low-income countries (LICs) is projected to increase by 187 percent in the same period. - For aging countries: pressures on fiscal sustainability and social cohesion; weaker growth due to declining labor force and total factor productivity (TFP) growth given a higher share of older workers. - For many LICs: potential demographic dividend that requires fostering opportunities for youth and avoiding capital shallowing. - Climate change: - Record-breaking extreme weather events are already affecting countries. - Natural disasters can disrupt output in the short term (workforce prevented from working, destroyed harvests) and in the medium term (physical capital destroyed and needing rebuilding). - The green transition poses significant challenges to firms, households, and policymakers, including for fossil fuel exporters. - Stranded assets and climate-related vulnerabilities may increase financial stability risks. - Figure metric: count of disasters with damages more than US$1 billion (source emphasis). - Geoeconomic fragmentation: - Defined as a policy-driven reversal of economic integration guided by countries’ strategic considerations. - Deepening fragmentation with a rising number of cross-border restrictions on trade and investment flows. - Expected consequences: erosion of gains from international trade and international risk sharing, weaker growth, and higher volatility. - Recent intensification of trade tensions and geopolitical conflicts likely to increase calls for energy security and self-sufficiency, strengthening incentives for inward-looking policies and further deepening fragmentation. - AI: - Recent advances could trigger a new wave of automation and profoundly change economic structure. - Productivity gains promise to boost growth, especially in advanced economies (AEs) where sectoral exposure to AI and digital preparedness are higher. - Magnitude and timing of gains are highly uncertain. - Rapid AI adoption could pose challenges for effective regulation, raise energy prices, and increase cybersecurity and financial stability risks. - AI could affect income and wealth inequality via job displacement and greater market concentration and power. - Figure metric: AI token usage by source type shows +3,700 percent growth for certain sources (weekly token volume in trillions). - Overall net effect on growth: - Net effect of these structural transformations on growth is estimated to be negative on average and for most economies. - Negative climate and geoeconomic fragmentation effects combine with unfavorable demographic trends in a wide range of economies, with only a partial offset from AI’s growth-enhancing effects. ### B. Implications for Macroeconomic Policy Frameworks - Context: - Existing policy frameworks (fiscal policy, monetary policy, macro-prudential policy measures (MPMs), foreign exchange intervention (FXI), and capital flow management measures (CFMs)) were largely developed during the Great Moderation and have been adapted post-global financial crisis and COVID-19. - Heightened uncertainty and more frequent/larger supply shocks may require further adjustments. - This paper focuses on macro policy frameworks; it does not address financial regulation or the role of broader structural reforms. - Pressures on policy frameworks: - Fiscal policy: - Public debt and deficits already elevated. - Sharper trade-offs between preserving fiscal space and output stabilization. - Pressures intensify if growth slows (for example, if potential AI productivity boosts fail to offset other drags) and spending rises because of demographic transition, climate change, climate transition, and geoeconomic fragmentation. - Fiscal rules, anchors, and reaction functions may need recalibration to maintain fiscal space and contain macroeconomic and financial stability risks associated with higher debt levels. - Monetary policy: - More frequent and persistent supply shocks complicate inflation–output stabilization trade-offs. - Long-term effects of structural transformations on natural interest rates (r*) are uncertain, but r* is generally expected to remain low. - Low r* suggests monetary policy frameworks may again be challenged by the effective lower bound (ELB), as before COVID-19. - External sector policies: - Increased capital flow and trade flow volatility may lead to greater use of FXI, CFMs, and MPMs. - Greater attention needed to integrated use of policy tools as guided by the IMF’s Integrated Policy Framework (IPF). - Need to revisit FX reserve adequacy frameworks and strengthen the global financial safety net (GFSN). - Calibration and policy mix: - Structural transformations may exert conflicting pressures on monetary, fiscal, and external sector policies, amplifying trade-offs. - Importance of a more holistic approach to managing available policy space. ### Questions Addressed by the Paper - The paper addresses three sets of questions: - (1) How will structural transformations and their interactions affect long-run growth? - (2) How will these structural transformations influence the nature of shocks/risks economies face, the associated transmission channels, and other key macroeconomic variables? - (3) What are the implications for macroeconomic policymaking? What adjustments are needed to ensure that fiscal, monetary, and external sector policy frameworks remain effective in maintaining domestic and external stability? *IMF DEPARTMENTAL PAPERS • Resilience amid Structural Transformations — 1. Introduction* ### 2. Economic Effects of Structural ### 2. Economic Effects of Structural Transformations ### Long‑Run Growth Impacts: approach and scope - Analysis covers 143 economies over the next decade (2026–35). - Estimated impact for each structural transformation = difference between annual real GDP per capita growth under the transformation and a hypothetical baseline that does not include the transformation going forward. - Exercise isolates moderate‑impact scenarios for each transformation without assigning likelihoods; estimates are not simple add‑ons to baseline forecasts because some effects may already be internalized by forecasters. - Methodology aims for parsimonious, consistent approaches across a large sample; specific methodologies differ by megatrend (see Annex I referenced in source). ### Methodologies by megatrend - Demographics: - Based on staff calculations using a Cobb‑Douglas production function and UN World Population Prospects Dataset 2024 vintage. - Incorporates a reduction of TFP in economies with larger shares of older workers (as in Aiyar and Ebeke 2016). - Moderate scenario uses median UN population projections by five‑year age groups; GDP depends on dynamics of population aged 15–65. - Uncertainty around population projections over the forecasting horizon primarily affects total population; working‑age population largely predetermined for horizons shorter than 15 years. - Demographic uncertainty around the moderate scenario: deviations of about +0.2 and −0.2 percentage points per year for low‑ and high‑fertility scenarios, respectively (total GDP unchanged over the period). - Migration flows taken at their baseline projected values. - Climate change: - Estimates based on IMF staff assessment of slow‑moving (chronic) climate effects across economies using a panel econometric model linking temperature increases to macro variables. - Moderate scenario assumes “incremental progress” where current policy trends broadly continue and emissions fall only slowly; global temperatures keep rising beyond Paris goals. - Chronic effects available for all 143 economies; additional acute (extreme weather) effects available for a subsample of 49 AEs and EMs (acute effects available for a subsample of 32 AEs and 17 EMs). - Global average level effect of chronic damages by 2100 estimated at a moderate 8 percent using this approach. - Literature range for total effects under “current policies” scenario: 1 percent to 25 percent. - Geoeconomic fragmentation: - Based on a standard quantitative multicountry multisector general equilibrium trade model with iceberg trade costs (scenarios adapted from Bolhuis, Chen, and Kett 2023). - Estimates do not account for negative effects of fragmentation of FDI and financial flows. - Average long‑term global impact of geoeconomic fragmentation: negative 2.1 percent of GDP. - Model‑based literature reports losses ranging from about 0.2 percent (mild) to roughly 7 percent (severe), with country‑level losses potentially higher (up to ~8–12 percent) when technological decoupling is included. - Artificial Intelligence (AI): - Growth impacts derived by extending Acemoglu’s (2024) task‑based framework across economies. - Occupation‑specific AI exposure weighted by country‑specific wage bill shares to gauge potential productivity gains; gains adjusted for cross‑country wage differences. - Lower AI preparedness (digital infrastructure, human capital, regulatory capacity) slows adoption and reduces medium‑term productivity gains. - Capital is expected to adjust in line with productivity gains over the medium term. - U.S. estimate in this exercise is somewhat larger than Acemoglu (2024), reflecting more optimistic productivity assumptions; results are toward the lower end of estimates discussed in the literature. ### Estimated average yearly effects on GDP per capita growth (moderate scenario, 2026–35) - Advanced economies (AEs), average effects: - Demographics: −0.5 percentage points - Climate change (including chronic and acute effects): −0.3 percentage points - Geoeconomic fragmentation: −0.2 percentage points - Artificial intelligence: +0.3 percentage points - Estimated total impact (including both chronic and acute climate effects): −0.7 percentage points - Context: average annual per capita growth during 2015–24 = 1.6 percent - Note: AI would need to add around 1 percentage point to annual growth for the average AE to offset the effects of other transformations. - Emerging markets (EMs), average effects: - Geoeconomic fragmentation: −0.2 percentage points - Climate change (including chronic and acute effects): −0.4 percentage points - Demographics: −0.1 percentage points (heterogeneous across countries; some may gain) - Artificial intelligence: +0.2 percentage points - Estimated total impact (including both chronic and acute climate effects): −0.5 percentage points - Context: average annual per capita growth during 2015–24 = 2.4 percent - Low‑income countries (LICs), average effects: - Geoeconomic fragmentation: −0.2 percentage points - Climate change: acute effects likely significant but hard to quantify for many LICs - Demographics: +0.4 percentage points (potential strong tailwind, conditional on productive employment and avoidance of capital shallowing) - AI: productivity gains could be modest in the medium term given limited AI preparedness, high employment shares in low AI‑exposure sectors, and relatively low labor costs - Unclear whether 2015–24 average annual per capita growth of 2.3 percent can be lifted or sustained going forward ### Interactions among megatrends and combined effects - Megatrends may reinforce or offset each other; key interaction channels include: - Geoeconomic fragmentation restricting migration, exacerbating demographic pressures in aging economies and trapping excess labor in younger populations. - Fragmentation weakening international knowledge transfer, slowing AI diffusion and associated productivity gains. - Geopolitical tensions disrupting supply of fuel and critical commodities, slowing AI advancement and diffusion. - Rapid AI advances potentially prompting export controls and trade barriers, intensifying fragmentation. - Indicative estimate of interaction effects (selected channels: higher barriers to trade and migration, slower AI diffusion): - Could shave off an additional 0.4 percentage points of annual per capita growth on average in the moderate scenario. - This would bring the total average effect of the combined megatrends (not including acute climate effects) to over half a percentage point. ### Uncertainty and comparability with literature - Estimates are broadly in line with the literature but are illustrative given substantial uncertainty around magnitudes and channels. - Comparison: McKibbin and Liu (2025) projections cited as broadly similar for demographics, somewhat more positive for AI, and somewhat more pessimistic for climate change (specific numbers reported in the source for 2026–35: AEs demographics −0.5 percent, AI +0.6 percent, climate −0.7 percent; EMs demographics −0.2 percent, AI 0.3 percent, climate −1 percent; LICs demographics +0.4 percent, AI 0.2 percent, climate −0.9 percent). ### Box 1 — Climate change: evidence and range of estimates - No consensus on how temperature increases affect GDP; modeling strategies yield markedly different estimates for the same warming trajectory. - Chronic effects (gradual temperature changes) versus acute effects (sudden, high‑impact events); chronic effects generally better understood but still variable. - The chronic damage estimates used here follow IMF How‑To Note (Centorrino, Massetti, and others 2025) and Kahn and others (2021), using panel regressions linking persistent temperature increases to output. - Key quantitative points cited: - Global average level effect of chronic damages by 2100 (this paper’s approach): 8 percent - Literature range for projected global GDP loss by 2100 under current policies (selected studies shown in source Box Table 1.1): - Nordhaus and Boyer (2000): Projected Effect of 1°C Incremental Global Temperature Increase = 1 percent; Projected Global GDP Loss by 2100 = 2 percent - Weitzman (2012): 1 percent; 3 percent - Tol (2014): 1 percent; 2 percent - Nordhaus (2014): 1 percent; 2 percent - Dietz and Stern (2015): 3 percent; 13 percent - Burke, Hsiang, and Miguel (2015): 10 percent; 14 percent - Howard and Sterner (2017): 4 percent; 8 percent - Kompas, Pham, and Che (2018): 1 percent; 2 percent - Kalkuhl and Wenz (2020): 3 percent; 5 percent - Howard and Sylvan (2020): 4 percent; 8 percent - Kahn and others (2021): 4 percent; 7–13 percent - Waidelich and others (2024): 5 percent; 8 percent - Bilal and Känzig (2024): 25 percent; 44 percent - Wide range reflects uncertainty in how temperature increases translate into economic damages rather than differences in climate scenarios. *IMF DEPARTMENTAL PAPERS • Resilience amid Structural Transformations (chapter: 2. Economic Effects of Structural Transformations)* ### Box 2. Potential Impact of Artificial Intelligence (AI) Adoption on Growth: What We ### Box 2. Potential Impact of Artificial Intelligence (AI) Adoption on Growth: What We Know So Far ### Estimates of AI effects on labor productivity and GDP - Goldman Sachs (Briggs and Kodani 2023) early estimate: labor productivity gains from AI of around 1.5 percentage points per year on average over the next 10 years for the United States; after considering additional factors, those gains translated to 0.4 percentage points additional medium-term GDP growth. - Aghion and Bunel (2024): boost of around 1.0 percentage point per year to labor productivity over the same horizon. - This paper’s moderate scenario (following Acemoglu framework but with larger gains): around 0.3 percentage points per year for the United States. - Acemoglu (2024): expects no more than 0.1 percentage points of annual labor productivity growth due to AI in the next 10 years. - Comparative context: moderate scenario in this paper lies between the high TFP scenario in Cerutti and others (2025) and the upside scenario explored in the 2025 October World Economic Outlook (IMF 2025h). ### Range of long-run GDP effects and confidence intervals (from Acemoglu 2024) - Near-term GDP effects generally below 2 percent by 2030. - Rising losses over time to around 3–6 percent by 2050. - About 7–10 percent by 2100. - 95 percent confidence intervals widen with the horizon: roughly 0–10 percent of GDP losses by mid-century and about 1–20 percent by 2100. ### Key drivers of divergent estimates - Task exposure estimates: Eloundou and others (2023) and Svanberg and others (2024) estimate AI could perform tasks accounting for about 20 percent of the aggregate US wage bill today. - Adoption and profitability assumptions: - Acemoglu assumption: only 23 percent of tasks exposed to AI can be profitably automated in the next decade. - Alternative views: adoption could be much broader—up to two-thirds of jobs exposed, of which up to 80 percent could see productivity gains from AI adoption. - Differences in assumed adoption speed, share of tasks automatable, and share of adopters explain most of the variation across studies. ### Uncertainty on productivity gains, reallocation, and aggregate effects - Limited early-case evidence: only a handful of case studies currently inform likely productivity gains from AI in the workplace; more research needed across occupations. - Reallocation frictions: Filippucci, Gal, and Schief (2024) model frictions to factor reallocation, which may lower aggregate gains, especially if productivity gains remain concentrated in a few sectors. - Distributional and demand effects: - Income inequality may rise, especially for entry-level jobs with low complementarity to AI (Cazzaniga and others 2024), potentially dampening aggregate demand. - Upside risks include new job creation or faster innovation through increased research and development productivity—Baily, Brynjolfsson, and Korinek (2023) argue that under this scenario, US GDP could double in 20 years. ### Heterogeneous impacts across country groups and sectors - Emerging markets (EMs) and low-income countries: - Sectoral composition in many EMs skewed toward less AI-exposed sectors, lowering potential productivity gains (Cazzaniga and others 2024). - Cerutti and others (2025): accounting for lower AI exposure and preparedness, low-income countries will only see half the gains of advanced economies. - Additional factor: varying labor costs—cheap labor in many economies may make automation with AI less profitable; aging economies with scarce labor may benefit more. - Over the next 10 years, additional growth from AI is likely to be only about half (EMs) and one-third (low-income countries) of the effect in the United States. - Implication relative to offsetting other mega-trends: - For the average advanced economy in the moderate scenario, offsetting negative effects from other megatrends would require gains from AI close to 1.0 percentage point per year. - For the average EM in the moderate scenario, this would require gains close to 0.7 percentage point per year. - Many LICs, especially in Africa, retain demographic dividends; for those with adverse demographics, AI seems insufficient to fully offset negative effects. ### Interaction with natural rates, term premia, and investment - Natural rates projections: - AI’s effect on the natural rate projected by applying elasticities of the natural rate to TFP growth and inequality (top 10 percent capture assumption for AI gains), using elasticities from Platzer and Peruffo (2022). - If widespread AI adoption substantially boosts productivity growth or AI-related investment surges, natural rates could be higher than baseline projections. - Melina and Villa (2025): under a strong AI-led investment surge broadly consistent with the ICT investment boom in the United States in the late 1990s, natural rates can rise by 0.7 percentage points, which would result in a US natural rate that is double the baseline natural rate in AEs. - Offsetting factors: - Increased inequality (if AI gains concentrate at the top) could offset upward pressure on natural rates. - Fiscal pressures and increased supply of government bonds could raise natural rates in some economies by reducing convenience yields (Schnabel 2025); higher sovereign bond supply effects may appear in higher term premia rather than short-term rates. ### Scenario dependence and structural uncertainty - Estimates depend on emissions paths, adaptation, technological change, and assumptions about adoption speeds and profitability. - Empirical relationships may not hold under higher warming; many models omit tipping points, financial spillovers, and interactions with structural shifts. - Focusing on near-term realized damages understates lock-in risks—the macroeconomic consequences of being committed to a high-warming path (tragedy of the horizon). - Overall assessment: climate-related GDP losses are likely to increase over time and may become substantial by mid-to-late century, but magnitude remains highly uncertain and skewed toward adverse outcomes. *Source: IMF staff calculations and referenced studies as presented in the source content.* ### 1. Natural Rate of Interest and Effects of ### 1. Natural Rate of Interest and Effects of Transformations in Selected Economies ### Natural rate developments - Natural rates in advanced economies (AEs) have increased somewhat since the COVID-19 pandemic but only to around prepandemic levels, which were considered historically low. - Multiple estimation approaches (semistructural econometric models, central bank projections, private-sector long-horizon forecasts, and financial-market–based measures adjusted for term premia) point to a pattern where the natural rate: - declined persistently over several decades, - bottomed out during the pandemic, - increased after that, but only to around the low levels prevailing just before 2020. - For the United States, a range of approaches yields consistent implications that the natural rate rose after 2024 but only to a low level similar to that prevailing just before the pandemic. - AI’s projected effect on the natural rate is based on projected AI effect on TFP growth and inequality, assuming that all AI benefits are captured by the top 10 percent of income distribution, using elasticities implied from Platzer and Peruffo (2022). ### Term premia and market implications - Since the global financial crisis, term premia have remained at historically low levels, but there are signs of a potential reversal. - Term premia have risen somewhat alongside the steepening of yield curves since early 2025. - A normalization of term premia to pre–global financial crisis levels could offset much of the benefit from low natural rates, thereby increasing borrowing costs for both sovereigns and the private sector. - Factors that could push term premia higher: - Heightened macroeconomic uncertainty and volatility, leading investors to demand greater compensation for holding longer-term securities. - Rising sovereign debt issuance—reflecting expanding fiscal spending needs—intensifying concerns about debt sustainability. - In panel 2, term premia are the average from the Adrian, Crump, and Moench (2013) and Christensen and Rudebusch (2012) models. ### Fiscal implications of transformations - Transformation-related fiscal spending pressures reflect additional outlays linked to preexisting commitments and discretionary policy choices to accommodate ongoing structural transformations. - Estimated additional spending needs: - Could amount to 3–4 percent of GDP annually over the next decade (2025–35) for the combined demands noted in the source. - These needs include spending on defense, climate adaptation, health and pensions, infrastructure, and human capital (with the source assuming zero additional spending on human capital or infrastructure for AEs in the specific estimates). - Sovereign debt pressures are likely to intensify as spending needs rise while interest-growth differentials could widen in some economies, reflecting weaker growth prospects. - Fiscal space (difference between maximum sustainable debt (MSD) and actual debt) will likely be negatively affected by ongoing structural transformations: - MSD levels will likely decline because of higher risk premia, lower convenience yields, and potentially weaker growth (and hence lower tax base). - Permanent increases in age-related, defense, and infrastructure spending imply lower sustainable primary balances unless accompanied by permanently higher tax revenues. - The paper estimates effects of structural transformations for 128 economies using growth-impact estimates and relevant methodologies. - Results show significant heterogeneity across economies. - In AEs, MSD levels are expected to decline significantly due to proximity to maximum tax capacity and the combined effect of increased expenditure and weaker potential growth driven by demographics and geoeconomic fragmentation. - In EMs and LICs, MSD levels could increase, subject to the ability to mobilize untapped tax revenue and realize growth potential: - Closing up to half of the tax gap in EMs and one-quarter in LICs—corresponding to median revenue gains of 2.6 percent and 2 percent of GDP annually, respectively—could preserve fiscal space of approximately 23 percent of GDP over the next decade, on average. - Important considerations: - The MSD is a gross public debt concept and does not account for governments’ assets; assessing fiscal space would benefit from a broader balance sheet perspective that considers net debt or public sector net worth. - Shifts in MSD can be substantial—on the order of 50 percent of GDP over a 10-year period for some countries—driven by high baseline MSD levels in AEs and model features that do not impose a practical lower-bound default threshold. - For LICs, concessional and blended financing from development partners could help meet large transformational needs without excessive reliance on domestic tax increases. - For fossil fuel importers with limited scope for further tax increases, energy subsidy reform can be a key instrument to rebuild fiscal space. _Italic: IMF DEPARTMENTAL PAPERS • Resilience amid Structural Transformations (excerpts provided in the source content)._ ### 1. Model-Based Fiscal Space with and without Tax ### 1. Model-Based Fiscal Space with and without Tax Adjustment ### Model setup and panels - Panel 1 compares 2024 debt levels with MSD at baseline and after transformations under two scenarios: - (1) no tax increases and (2) optimal tax adjustment. - MSD is bounded at zero. - Optimal tax increases are capped relative to current forecast, up to half of the tax gap in EMs and one-quarter in LICs, corresponding to median revenue gains of 2.6 percent and 2 percent of GDP annually, moving closer to tax capacity, whereas AEs are assumed to not be able to increase revenues. - Panel 2 shows the ratios of primary balance (PB) and MSD (adjusted for the effects of megatrends, not including acute climate effects) relative to the current projections across income groups after 10 years. - Missing spending data are treated as zero. - Defense spending is based on 3.5 percent target for NATO countries. - The sample includes 128 economies. - Definitions and abbreviations preserved as in source: - AE = advanced economy; EMs = emerging markets; LIC = low-income country; MSD = maximum sustainable debt; NATO = North Atlantic Treaty Organization; PB = primary balance. ### Megatrends, tax capacity, and channels - The MSD analysis acknowledges additional factors not fully captured by the model: - Political and social constraints can limit revenue mobilization. - Structural transformations may reshape the tax base: Digitization and AI may improve VAT and personal income tax compliance; the energy transition could decrease carbon-related revenues. - Aging, automation, climate shocks, and fragmentation may erode revenues from labor taxes, corporate profits, and customs duties. - Generative AI could erode the labor income tax base and increase the importance of taxing capital income from data, algorithms, and intangible assets. - In the MSD analysis, AI’s effect occurs only through r−g; other channels (AI’s effect on relative importance of labor versus capital tax bases, labor displacement, direct public spending on the AI ecosystem) are not explicitly modeled. - Public infrastructure enabling AI (roads, electricity) is captured only indirectly. ### General equilibrium model and scenarios - The paper uses a general equilibrium model (Nguyen, Lam, and Sher 2025) extending existing frameworks, featuring: - A rich fiscal block with public investment and public consumption. - Multiple tax instruments. - Endogenous risk of debt distress increasing nonlinearly with the debt-to-GDP ratio. - Calibration to representative AE, EM, and LIC structural characteristics (representative AE proxied by GDP-weighted average across AEs, excluding the United States; EMs and LICs represented by the median country). - Three financing scenarios are considered: - Scenario A: all new spending financed through borrowing (full debt financing). - Scenario B: all new spending financed by adjusting government consumption (considered only for AEs). Example parameter: resulting in an increase of 1.2 percentage points of GDP out of 2.7 percent in baseline per year on average. - Scenario C: financing comes from expanding tax capacity (considered for EMs and LICs). Example parameters: - EMs: resulting in a total revenue increase of 2.4 percent of GDP per year on average over the next decade. - LICs: resulting in a total revenue increase by 2.8 percent of GDP per year on average over the next decade, with 80 percent of increases coming from lump-sum tax and 20 percent from consumption tax. - Public investment efficiency assumed at 0.85 (or 0.15 efficiency gap) across all country groups in model simulations. ### Simulation results and key quantitative findings - For a stylized AE: - Additional spending associated with four structural transformations amounts to roughly 3 percent of GDP per year. - If fully debt financed (scenario A), public debt would exceed the MSD threshold. - Most AEs already operate close to their tax capacity; accommodating new spending requires reallocating existing expenditures toward high-impact, growth-enhancing areas (scenario B), and potentially entitlement reforms. - For stylized EMs and LICs: - Additional expenditure pressures can be absorbed through a combination of debt financing and purposeful domestic revenue mobilization. - Many EMs and LICs operate with tax-to-GDP ratios persistently below 15 percent. - Scenario C findings: - Increasing tax revenues by an average of about 2.8 percent of GDP in LICs and 2.4 percent of GDP in EMs per year over the next decade enables financing additional spending while keeping debt comfortably below MSD thresholds and maintaining adequate fiscal buffers. - Buffer and stock-flow context: - The buffer is estimated based on Eyraud and others (2018) using the 1981–2025 sample to keep debt under threshold with a high probability (90 percent). - For LICs, average stock-flow adjustment over 2000–25 is about 1–1.5 percent of GDP per year on average; a buffer of 10 percent of GDP is in this range. - Spending efficiency caveat: - Lower spending efficiency would lead to a higher debt increase because of smaller effects on GDP for each dollar spent. ### Policy recommendations and possible adjustments - Prioritize Debt Sustainability - Pivot fiscal priorities toward mitigating sovereign risks and establishing credible MTFFs. - Adopt a risk-based approach centered on a medium-term debt anchor and explicit safety margins to keep debt below critical limits with high probability. - Reassess fiscal anchors intermittently—every four to five years—and consider approaches that capture a broad range of risks. - Incorporate robust control principles into fiscal framework design given high uncertainty surrounding megatrends. - Recalibrate Fiscal Rules and Reform Entitlements - Fiscal rules should be transparent, comprehensive, and sufficiently robust to remain binding under stress; expenditures linked to structural changes should remain within the scope of fiscal rules. - Regular stress testing of fiscal frameworks against adverse scenarios (prolonged low growth, higher interest rates, contingent liabilities) is critical. - Strong, well-designed error-correction mechanisms with clear triggers and timelines are necessary to guide policy back to anchors while balancing decisiveness and gradualism. - Comprehensive long-term fiscal sustainability analysis and timely entitlement reforms are needed to address rising age-related health and pension spending. - Strengthen Fiscal Management - Reduce spending rigidities arising from legally mandated expenditures, entrenched entitlement programs, or inflexible budget processes; embed systematic reviews in medium-term budget frameworks. - Improve spending efficiency through transparent procurement and robust public investment management. - Strengthen fiscal institutions, particularly fiscal councils, ensuring operational independence, clear statutory mandates, adequate resources, and direct public communication channels. - Redesign Fiscal Instruments and Upgrade Social Safety Nets - Adjust fiscal systems to longer working lives and slower labor force growth: support lifelong learning, reskilling, age-friendly workplaces, flexible work arrangements, and remove age-based disincentives in tax and benefit systems. - For AI-driven transformations, rebalance the tax system away from labor and toward capital and corporate income; redesign automatic stabilizers and social safety nets to address structural labor market displacement. - Upgrade social protection in dimensions of coverage, duration, interaction with reskilling programs, and consider instruments such as wage insurance. - For climate and energy transitions, redesign fiscal frameworks for fossil fuel exporters to manage declining hydrocarbon revenues and stranded assets and fiscal risks from state-owned enterprises in carbon-intensive sectors. - Pay Greater Attention to Fiscal–Monetary Policy Interactions - Recognize limited stabilization role for fiscal policy in large supply shocks when debt is high; focus on temporary and targeted interventions. - Improve shock identification and coordination to avoid conflicts between fiscal and monetary policy as supply shocks can generate demand-side spillovers. - Adopt More Proactive Financing Strategies - Integrate fiscal frameworks with debt-workout mechanisms and state-contingent instruments (including disaster clauses) to enhance risk sharing and smooth debt service. - For EMs and LICs: strengthen public debt management and domestic debt market development through resilient debt structures, stronger DMO institutions, debt transparency, active investor relations, and local-currency bond issuance. - For economies near tax capacity, embed the IMF–World Bank three-pillar approach: domestic reforms first, external financing support second, complemented by measures to reduce effective debt service burdens and mobilize private capital. *Source: IMF staff estimates.* ### 4. Monetary Policy Frameworks ### 4. Monetary Policy Frameworks ### A. Challenges - Structural transformations affecting potential GDP growth, the natural rate of interest, and the nature of shocks raise questions about fitness of existing monetary policy frameworks, with focus here on inflation-targeting (IT) frameworks. - Decomposition of core inflation (sign-restricted VAR on cyclical GDP and core CPI with stiff HP filters) shows both supply and demand shocks contributed significantly in the postpandemic recovery and after the war in Ukraine, with the contribution of supply shocks particularly elevated across AEs and some EMs; data are reported as quarter-on-quarter annualized inflation. - Five main challenges for IT regimes: - (1) More frequent supply shocks and sharper inflation–output trade-off. Negative supply shocks raise inflation while suppressing output, complicating policy trade-offs relative to demand-driven fluctuations. - (2) Greater uncertainty. Increased frequency/severity of shocks and changes in economic responses reduce the effectiveness of relying on a single baseline forecast and complicate communication and credibility. - (3) Increased risk of fiscal dominance. High fiscal vulnerabilities may constrain scope for monetary tightening; fiscal dominance can erode purchasing power and lead to financial repression, undermining central bank ability to meet inflation targets. - (4) Continued ELB risk. Baseline estimates suggest natural rates are likely to remain low, implying some central banks may again confront the ELB and be constrained in supporting demand via conventional tools. - (5) Changes in monetary policy transmission mechanisms. Population aging, climate change, and geoeconomic fragmentation are expected to weaken transmission; AI effects are uncertain and could either enhance price flexibility or reduce interest sensitivity and alter financial-sector dynamics. - Additional nuances: - EMDEs face more volatile macroeconomic conditions and more volatile inflation due to larger food and energy CPI weights, stronger shock amplification, and weaker anchoring of inflation expectations, increasing trade-offs and placing a premium on central bank credibility. - When mark-up supply-side shocks predominate, standard sign-restricted VAR decompositions can misclassify shocks, but the decomposition results here align with alternative methods for the United States. ### B. Possible Adjustments - IT frameworks have shown resilience historically, suggesting incremental adjustments rather than wholesale redesign; however, many adjustment options involve trade-offs, implementation challenges, and credibility risks. - Four key priorities for refining IT frameworks: - (1) Determining conditions under which policymakers can “look through” supply shocks. - Historically, many AE IT central banks looked through transient supply shocks when medium-term inflation expectations were unlikely to be affected; in EM and LIC IT countries “looking through” is more limited by credibility concerns. - When supply shocks are large and persistent, monetary policy may need to respond more forcefully to prevent shocks being embedded in inflation expectations. - Appropriate response depends on state of the economy, level of inflation, and degree of central bank credibility; when inflation is already elevated, “looking through” may risk de-anchoring expectations. - Cyclical position matters: shocks during an economy running above or close to potential amplify pass-through and wage responses. - Further work is needed to clarify when looking through remains appropriate and how structural transformations may change interest sensitivity and the Phillips curve. - (2) Making more systematic use of alternative scenarios to reflect greater uncertainty and to inform policy decisions beyond a single baseline forecast. - (3) Safeguarding central bank independence to mitigate fiscal dominance risks and preserve the ability to achieve inflation objectives. - (4) Reviewing the policy toolkit at the ELB, including consideration of nonconventional tools and the conditions for their use. - Cautions about specific adjustments: - Raising the inflation target (unless the current target is very low, that is, below 2 percent) could undermine credibility. - Shifting from a fixed target horizon to a more flexible horizon may help with trade-offs but entails credibility risks if independence and credibility are not firmly established. - Reviewing tolerance band widths should be approached with caution. - Annex 5 (Annex Table 5.1) lays out pros and cons of other potential adjustments. ### Box: Monetary Policy Implications of Elevated Shock Uncertainty (Box 4) - Simulation context: IMF’s Quantitative Integrated Policy Framework (QIPF) model augmented with endogenous wage indexation. - Key simulation insight: When inflation is more elevated to begin with, repeated adverse supply shocks can produce a more pronounced adverse inflation cycle; stabilizing inflation after repeated shocks may require larger policy rate responses. - Illustration: An initial global cost-push supply shock at t = 1 is followed by an identical shock hitting at t = 4, 6, or 8 quarters (blue, red, yellow lines). For the second shock at 4 quarters (blue), the inflation response is only marginally larger than to the first shock, but the policy rate hike from t = 4 onward must be significantly larger than the rate increase following the first shock at t = 1 to achieve that outcome. *Source: 4. Monetary Policy Frameworks, rastea - 4. Monetary Policy Frameworks (PDF chapter).* ### 3. Output4. Government Debt (Percent of GDP) ### 3. Output4. Government Debt (Percent of GDP) ### Simulation findings (Impulse response functions, QIPF) - Model and data: - Impulse response functions based on the IMF’s QIPF model. QIPF = Quantitative Integrated Policy Framework. - APR = Annualized Percentage Rate. - Key results: - Counteracting inflation risks from already elevated wage indexation, even if the second shock is identical to the first, requires a much larger monetary policy response, with significant additional output costs (panel 3). - When the second shock hits later, the red and yellow lines depict a similar effect, although the risks recede as inflation gets closer to target before realization of the second shock. - Repeated shocks may give rise to nonlinearities that increase the risk of inflation de-anchoring if authorities do not tighten monetary policy sufficiently and in a timely manner. - Panel 4 shows how surprise inflation may improve the government debt position in the near term although output and tax revenues fall. - Source note: - Source: IMF staff calculations. ### Make more systematic use of alternative scenarios under increased uncertainty - Rationale: - Diagnosis of the nature, persistence, and transmission of shocks and model uncertainty will likely remain nontrivial even as real-time shock identification improves. - A structured, small set of internally consistent alternative scenarios alongside the baseline forecast would support a risk-management approach to monetary policy. - Recommended practices: - Routinely explore a set of internally consistent narratives and stress-test policy choices in these scenarios (Bank of England 2024). - Keep the scenario set small to avoid overcomplicating deliberations. - Use scenario analysis to: - Assess robustness of decisions under different assumptions about exogenous variables (such as energy prices) or the structure of the economy. - Highlight potential nonlinearities and tail risks. - Clarify how policy trade-offs might shift in alternative states of the world. - Support a more gradual and flexible policy approach, allowing policymakers to adjust stance as new information arrives. - Future research/priorities: - Improve capacity to distinguish between transitory and persistent shocks in real time. - Strengthen scenario design and develop frameworks for constructing alternative narratives and conveying policy implications. - Promising directions from recent research: - Combining statistical models with judgment-based scenarios to integrate diverse risks into policy forecasts (Adrian, Giannone, and others 2025). - Probabilistic weighting approaches that update likelihoods of the baseline and limited alternatives as new information arrives (Cairó and others 2025). - Acknowledge continued challenges when models yield conflicting signals or when shocks fall outside historical experience. - Communication: - Communicate uncertainty through scenario analysis as part of central bank communications to help the public understand salient risks and the central bank’s reaction function. - Principles for scenario communication: - Establish a hierarchy of scenarios. - Standardize key metrics. - Offer concise explanations of underlying assumptions. - Avoid excessively detailed scenario presentations that could overwhelm the public and obscure key messages. ### Strengthen central bank independence and mitigate fiscal dominance risk - Key proposition: - Mitigating fiscal dominance ultimately hinges on reducing fiscal vulnerabilities and preserving or strengthening central bank independence. - Elements to reinforce: - Operational autonomy: full control over monetary policy instruments and prohibition of monetary financing. - Institutional and personal independence: transparent appointments, fixed and staggered terms, narrowly defined conditions for dismissal. - Financial independence: adequate capitalization, predictable recapitalization arrangements, and autonomous budgeting (Adrian, Khan, and Menands 2024; IMF Safeguards Assessment). - Accountability and transparency frameworks, including adherence to the IMF’s Central Bank Transparency Code (IMF 2020). - Additional considerations: - In some IT regimes, more commonly in EMDEs, strengthen the inflation target as the nominal anchor. - Reinforce commitment to exchange rate flexibility to reduce inconsistent policy signals and enhance credibility (Ötker-Robe and Vavra 2007). - Consider a medium-term review cycle for the inflation target (for example, a five-year review cycle) to limit ad hoc adjustments and support anchored expectations (Carter, Kamber, and Otten 2026). ### Review the policy toolkit at the effective lower bound (ELB) - Context: - Real interest rates expected to stay low on average, reflecting persistently low natural rates in AEs and a likely gradual convergence toward lower natural rates in EMs, with occasional supply-driven inflation and rate surges. - Risk of exhausting conventional monetary policy space remains in AEs and may emerge in EMs, especially under negative demand shocks. - Unconventional monetary policy (UMP) toolkit recommendations and considerations: - Negative interest rate policy: - May play a greater role; experience shows the ELB lies lower than zero (example: Switzerland lowered its policy rate to −0.75 percent). - Negative rates can provide meaningful stimulus (Brandão-Marques and others 2024; Kolasa, Lassen, and Lindé 2025). - Quantitative easing (QE): - Important for prolonged liquidity traps; use should typically be reserved for such episodes and governed by clearly defined exit clauses (Adrian, Erceg, and others 2025). - QE can boost private demand in major economies and provide additional stimulus in open economies via the exchange rate channel. - QE carries risks, especially in countries with weaker institutions where distinctions between QE and monetary financing may blur (Hofman and Kamber 2020); large balance sheets can generate losses complicating fiscal–monetary relations (Adrian, Erceg, and others 2024). - Need for strong safeguards and transparent communication around activation and unwinding of QE. - Given the convexity of the Phillips curve, likely greater frequency of large supply shocks, and unwinding challenges, a greater tolerance for modest inflation undershooting may be warranted when long-term inflation expectations remain well anchored. - Funding-for-lending (FFL) programs: - Can strengthen monetary transmission by lowering bank funding costs and encouraging credit expansion when rate cuts lose traction. - Risks include potential credit misallocation, financial stability pressures, and challenging exit dynamics; net benefits viewed as positive in stressed environments but diminish as policy rates normalize (BIS 2023). - Forward guidance: - Shapes expectations about the future policy path and influences longer-term interest rates. - Effectiveness depends on credibility and clarity; risks locking in suboptimal policies when combined with other UMP measures. - Deploy with well-designed escape clauses. - Yield curve control (YCC): - Mixed results in Australia and Japan; can support transmission under certain conditions but experienced impaired market functioning and exit difficulties, raising credibility concerns (RBA 2022; IMF 2023e). - Central banks should consider QE ahead of YCC in the future. - Open questions and research needs: - Timing and calibration of QE, appropriate tolerance for temporary undershooting of the inflation target, and design of escape clauses. - Structuring FFL programs to support lending while minimizing credit misallocation and financial stability risks; determining exit strategies that avoid abrupt tightening in bank funding conditions. - Assess FFL effectiveness relative to other UMP tools. - For EMDEs, decisions on UMP tools must account for risks to central bank credibility linked to perceptions of fiscal dominance. - Figure 11 summary (Challenges and Possible Adjustments) — thematic points preserved: - Challenges: More frequent supply shocks; Higher uncertainty about the nature and transmission of shocks; Fiscal dominance risk; Effective lower bound risk. - Possible Adjustments: React more decisively to large supply shocks and establish conditions when “looking through” remains appropriate; Analyze how megatrends affect the inflation-output trade-off; Make greater use of risk scenarios to formulate and communicate uncertainty and the central bank reaction function; Preserve/strengthen central bank independence; Pay greater attention to effective fiscal and monetary policy interactions; Consider more use of negative interest rates, forward guidance with escape clauses, cautious use of QE, tolerate small inflation undershooting. - Notes: - QE = quantitative easing. - Structural features of some financial systems—such as the relatively large size of the money market fund sector—may make some policymakers reluctant to deploy negative policy rates in the future. - QE here refers to purchases undertaken for macroeconomic stimulus rather than for financial stability purposes. - Any unconventional monetary actions should depend on the gains and perceived distribution of risks attributed to these tools. *IMF DEPARTMENTAL PAPERS • Resilience amid Structural Transformations* ### 5. External Sector Policy Frameworks ### 5. External Sector Policy Frameworks ### IPF-based approach and deployment principles - The IMF’s Integrated Policy Framework (IPF) provides a systematic framework for the use of additional policy instruments—FXI, CFMs, and MPMs—alongside fiscal and monetary policy tools to help maintain macroeconomic and financial stability. - Central IPF principle: FXI, CFMs, and MPMs should be deployed only when specific frictions justify their use and when benefits outweigh costs. - Exchange rate flexibility is generally most beneficial in countries with floating exchange rates and limited financial frictions; additional tools may be warranted in the presence of frictions. - Distinction between shocks: - Fundamental shocks: require macroeconomic adjustments to restore stability. - Nonfundamental shocks: arise from market imperfections (e.g., “noise trader shocks”) and are generally temporary. ### IPF use cases for additional instruments - Use case A (arbitrage frictions in shallow FX markets) - Volatile capital flows can result in destabilizing financing premia. - Addressing nonfundamental shocks through FXI and/or CFMs can give monetary and fiscal policy greater flexibility, provided use is consistent with IPF principles. - Use case B (FX mismatches and financial amplification) - Large depreciations can increase financial stability risks when there are sizable FX mismatches on balance sheets of households, nonfinancial corporates, and financial sectors. - FXI to counter financial amplification should be limited to smoothing severe depreciations. - Use case C (de-anchoring of inflation expectations) - Large exchange rate movements can de-anchor inflation expectations and pose risks to price stability. - FXI can support monetary policy when a large depreciation threatens to de-anchor expectations, but only when: - reserves are sufficient, - the marginal cost of relying solely on monetary policy is high, - and the costs of deploying FXI are low. - FXI may also play a role in leaning against sustained appreciation pressures in a low inflation and low rates environment. ### A. Challenges arising from structural transformations - (1) More volatile trade and capital flows - Geoeconomic fragmentation can cause more changes in tariffs and nontariff barriers → more volatile trade flows and more frequent terms-of-trade shocks. - Trade restrictions on high-tech goods and the AI race could further realign global supply chains. - Reconfiguration of capital flows could make it more difficult for vulnerable EMs and LICs to meet external financing needs amid rising adaptation and transformation spending pressures. - (2) Changing shock characteristics and higher uncertainty about shocks - Structural transformations likely increase frequency and severity of both fundamental and nonfundamental shocks. - Higher uncertainty increases likelihood of risk-off episodes triggered by fundamental shocks (example: AI-powered algorithmic trading could amplify market volatility). - Breakdown in historical relationships between the US$ and other macrofinancial variables could lead to abrupt shifts in investor behavior. - (3) Deepening financial frictions that heighten vulnerability to financing shocks - FX market depth may decline due to restrictions on cross-border financial flows and reduced foreign investor participation. - Balance sheet mismatches: FX mismatches could rise if investors take on less duration risk or borrowers shift toward non-US$ foreign currency liabilities. - Higher aggregate uncertainty can strain FX markets by raising funding costs, reducing liquidity, and amplifying excess exchange rate volatility; FX market stress can spill over to other asset classes, posing macrofinancial stability risks—particularly where currency mismatches and fiscal vulnerabilities are significant. - (4) Availability and effectiveness of policy instruments and reserve buffers - CFMs: Higher exchange rate volatility may increase incentives to use CFMs; expansion of crypto assets (including stablecoins) could facilitate CFM evasion and weaken enforceability of residency-based CFMs. - FXI and FX reserves: - FXI may become more frequent, potentially requiring higher FX reserves. - Reserve adequacy remains uneven; some countries may require higher reserve buffers in a more fragmented global financial system with fewer risk-sharing opportunities and a higher incidence or severity of shocks. - Illustration from sample of 40 EMs: about half the sample had adequate reserves in 2025 according to the IMF’s Assessing Reserve Adequacy (ARA) metric; when the probability of sudden stops rises to historical levels associated with high global risk aversion, the share of EMs with reserve shortfalls increases from under one-half to two-thirds of the sample. - Simulation detail: right-hand bar in Figure 12 shows needed increase in reserves/ARA if probability of a sudden stop increases from 10 percent to 20 percent. - Reserve adequacy guidance: reserves in the range of 100–150 percent of the metric are considered to be adequate. ### B. Possible adjustments and policy priorities - Reduce vulnerabilities to increase resilience to external shocks - Prioritize ex ante measures to reduce vulnerabilities, including use of MPMs and broader efforts to lessen financial frictions. - Diversification of economic and financial exposures (including trade patterns and financing flows such as FDI) is important. - Use of MPMs can mitigate shock impact, allow for more decisive monetary policy, and reduce need for CFMs or FXI. - Example: borrower-based MPMs that contain financial stability impact of higher interest rates allow central banks to hike rates with less concern about defaults among local currency borrowers. - MPMs that limit FX mismatches on borrowers’ balance sheets can reduce need for FXI under use case B. - Preserve exchange rate flexibility - For countries with floating regimes, exchange rate flexibility remains important to facilitate adjustment to fundamental shocks. - For countries facing significant financial frictions, FXI remains a useful part of the toolkit. - Higher incidence of shocks raises sharper intertemporal trade-offs between deploying reserves now versus accumulating reserves for future shocks and raises the threshold for using limited FX reserves. - More attention likely to be devoted to currency composition of reserves in fragmented settings. - Maintain adequate buffers and strengthen the Global Financial Safety Net (GFSN) - Greater volatility in trade and capital flows implies implications for level and composition of adequate FX reserves; scenario analysis becomes more important in assessing reserve adequacy. - Strengthening the GFSN requires addressing uneven access, enhancing crisis preparedness, and better coordination across GFSN layers. - Increased demand for safe assets may raise the cost of holding reserves, suggesting revisiting the role and design of reserve pooling arrangements and exploring ways to reduce the cost of building reserve buffers globally. - Priority areas for further analysis and operational work - Develop empirical toolkit for real-time identification of shocks and drivers of exchange rate movements to help identify episodes that may warrant policy intervention. - Improve monitoring of CFMs via enhanced data collection and analytical capabilities to track digital money flows and assess macroeconomic effects. - Assess feasibility of moving from residency-based CFM design toward transaction-based or risk-based approaches. - Provide additional guidance on deploying FX reserves most effectively, including design of targeted FXI modalities. - Upgrade reserve adequacy framework to account for higher volatility in trade and capital flows and mismatches between currency composition of external liabilities and FX reserves; help countries calibrate precautionary reserves and determine pace of buffer rebuilding following shocks. - ARA metric detail (for EMs with floating exchange rates) - ARA Metric = 0.05 × Exports + 0.05 × Broad Money + 0.3 × Short-Term Debt + 0.15 × Other Liabilities. - Adjustments to the Assessing Reserve Adequacy metric are made for countries with CFMs or fixed exchange rates. - Reserves in the range of 100–150 percent of the metric are considered to be adequate. ### 6. Considerations for non-IT frameworks - Prevalence and vulnerability - Although most AEs and many EMs employ IT frameworks alongside floating exchange rates, about 40 percent of EMs and LICs rely on the exchange rate as the nominal anchor, and a small share have monetary aggregate targeting frameworks. - About one-third of EMs and LICs have intermediate exchange rate regimes. - Risks and policy priorities for conventional pegs and soft pegs - Conventional pegs: larger or more frequent shocks place disproportionate burden on other policies—especially fiscal policy—given absence of a monetary policy lever; countries should prioritize building fiscal buffers and FX reserves and could consider increasing exchange rate flexibility over time to develop independent monetary policy, provided other country-specific factors support the transition. - Intermediate arrangements: focus on clarifying framework and prioritization of policy objectives and on effective operational design. - Soft peg with exchange rate band: authorities may be tempted to widen the band to absorb larger or more frequent shocks, but bandwidth should reflect country-specific conditions and should not be so wide as to undermine anchor credibility or nominal stabilization role. - Crawling pegs: adjustments to the rate of crawl should reflect structural transformations and associated shocks to home and foreign economies, but reviews should be infrequent to preserve credibility; contingency arrangements for off-cycle reviews may be desirable for exceptional circumstances (e.g., unexpected large capital flows). - Operational and capacity constraints - Non-IT EM and LIC central banks with scope for independent monetary policy could adopt scenario analysis and strengthen central bank independence to counter fiscal dominance risk; where capacity is low, strengthening core operational capacity may need to precede more sophisticated risk-based approaches. - Principles for use of additional instruments in non-IT settings - FXI, CFMs, and MPMs should be used in ways that strengthen, not undermine, commitment to the nominal anchor and to macroeconomic and financial stability. - Calibration of policy responses must account for common challenges—particularly in LICs—including low reserves, weak monetary transmission, limited operational capacity, and governance constraints. - FXI, CFMs, and MPMs should support rather than substitute for necessary monetary and fiscal adjustments; inappropriate deployment that offsets needed macroeconomic tightening can blur objectives, diminish effectiveness, and erode credibility. - Policies should take a medium-term perspective and focus on building stronger frameworks, including enhancing fiscal sustainability to reduce fiscal dominance risks and gradually reducing financial frictions over time. *Source: IMF Departmental Paper — “Resilience amid Structural Transformations,” Chapter 5: External Sector Policy Frameworks.* ### 7. Policy Interactions and Trade-Offs ### 7. Policy Interactions and Trade-Offs ### Monetary–Fiscal Interactions: Overview and Mechanisms - Structural transformations make shocks larger, more persistent, and more frequent, while rising spending pressures (demographics, climate change, geoeconomic fragmentation) may significantly constrain fiscal space. - Nonlinear price and wage Phillips curves and endogenous indexation imply state-dependent inflation responses: when inflation exceeds the central bank’s target persistently, wages and prices become increasingly determined by lagged inflation, producing disproportionately larger and more persistent inflation responses. - Model setup: variant of the quantitative micro-founded IPF model (Adrian and others 2021), two-country open-economy New Keynesian DSGE with nonlinear Phillips curves and endogenous indexation; de-anchoring of inflation expectations following large adverse supply or positive demand shocks (as in Erceg, Lindé, and Trabandt 2024). - Policy intuition for a temporary global supply (cost-push) shock: - Standard-sized shocks: monetary tightening alone generally sufficient to stabilize inflation. - Large shocks: nonlinearities and indexation increase inflation persistence; monetary policy alone would require much larger interest rate hikes, producing larger declines in output and private consumption and increases in government debt. - Fiscal responses through automatic stabilizers and targeted transfers can mitigate output losses and protect vulnerable households but may raise inflation risks; targeted design limits those risks. - Fiscal consolidation that offsets increased government debt can allow central banks to tighten less aggressively while achieving similar inflation reductions. ### Scenario-Based Findings (A–E) — Key Results - Scenario A (typical AE, deep FX markets, without foreign debt): - A tighter fiscal stance can deliver similar output and inflation outcomes as greater monetary tightening, but with lower private consumption losses and lower debt because of reduced expenditures and lower interest rates. - Fiscal policy playing a larger stabilization role is more compelling as evidence accumulates on nonlinear inflation responses. - Driven by adverse supply shocks, fiscal policy that aims to stabilize debt typically implies a procyclical stance; if strong demand contributes to inflation, the fiscal stance could remain countercyclical. - Scenario B (vulnerable economy with high foreign debt, significant negative NFA, and shallow FX markets): - Fiscal consolidation leaves the real exchange rate unchanged because of offsetting interest rate differentials and the improvement in the NFA position. - Fiscal consolidation is even more effective at curbing inflationary pressures than in the nonvulnerable case. - Expansionary fiscal policy would be counterproductive: it would force central banks to lean against the stimulus and ultimately leave the economy more indebted. - Scenario C (repeated moderate shocks — Box 4): - Repeated moderate shocks arriving before inflation is fully stabilized can generate dynamics similar to a single large shock. - Elevated inflation triggers nonlinearities even when single shocks hitting closer to target would have triggered only modest nonlinearities. - Stabilizing inflation then requires a much larger secondary monetary policy response, with larger output and private consumption losses. - Faster adjustment when inflation has risen significantly above target plays an important role when shocks are large or frequent. - Scenario D (large cost-push shock when monetary and fiscal frameworks are weak): - In a country with weak frameworks, much more aggressive monetary tightening is required, yet inflation remains elevated and debt rises sharply. - When fiscal space is limited, economies with weaker frameworks may still need to prevent further debt deterioration. - Scenario E (extension of D: require government debt to return to pre-shock level within five years): - Different combinations of monetary and fiscal tightening produce different growth and inflation outcomes. - Fiscal authorities need to consider joint impact of monetary and fiscal adjustments, accounting for the central bank’s reaction function. - Improving policy frameworks before shocks hit is crucial to avoid more painful trade-offs as structural transformations intensify. Notes on model calibration and shocks: - Example numerical setup: response to a large 3 standard deviation shock at time t = 1 (Figure 15 illustration). - Time-horizons referenced for framework credibility: central banks keeping inflation forecast and expectations close to target at the end of the forecast horizon (three years); bringing government debt back to preshock value within five years. ### Cross-Policy Trade-Offs Amplified by Structural Transformations - Fiscal constraints limiting monetary and external policy: - Elevated fiscal risks can raise financing premia, induce capital flight, and complicate macro management. - Unsustainable fiscal paths weaken monetary transmission through expectations, risk premia, and exchange rate (Wu and Xie 2025). - Higher term premia on government debt may translate into higher private borrowing costs, forcing monetary offset and reducing policy space. - Monetary policy constraints limiting fiscal responses: - When monetary policy is at the ELB and public debt is near sustainability limits, fiscal stimulus may not be feasible even after a negative demand shock (Adrian, Erceg, and others 2025). - Under fiscal dominance, high policy rates and rising interest expenses may limit flexibility to provide fiscal stimulus (IMF April 2025 Fiscal Monitor). - External sector constraints limiting fiscal and monetary options: - Need for higher FX reserve buffers may raise the bar for using FX intervention (FXI) in response to external shocks. - The IPF model shows FXI can improve policy trade-offs in economies with nominal frictions where depreciations are contractionary; insufficient reserves may force monetary or fiscal tightening after financing shocks or depreciations with FX mismatches. - Financial stability constraints on monetary policy: - Fragmentation-driven higher cross-border financial barriers reduce international risk sharing and increase risk premia. - Rapid AI equity valuations or climate-related stranded asset risks may heighten financial stability concerns, limiting scope for higher policy rates. - Robust prudential frameworks can mitigate these risks and grant monetary policymakers greater flexibility. ### Policy Recommendations and Design Implications - Strengthen policy frameworks ex ante: - Improving monetary and fiscal frameworks before shocks hit reduces need for painful trade-offs later. - Maintain central bank independence: operational control over policy instruments, prohibition of monetary financing, and strong institutional, financial, and transparency frameworks. - Make fiscal anchors risk based and state-contingent: - Static rules calibrated to past volatility may be inadequate in a more shock-prone world. - Design risk-based fiscal anchors that balance credibility with flexibility when shocks are more frequent and persistent. - Calibrate fiscal instruments and financing mixes: - Debt-financed responses quickly run up against debt sustainability constraints, especially in AEs. - Revenue mobilization can alleviate constraints in most EMs and LICs; further work needed on political economy, distributional consequences, and growth effects of fiscal instruments. - Combine spending cuts with modest targeted transfers to vulnerable households to mitigate consumption inequality impacts of fiscal tightening. - Operationalize “looking through” supply shocks in IT frameworks: - Larger, more persistent shocks interacting with wage setting and expectations formation complicate the decision to “look through.” - Systematic use of scenarios, stress tests, and sensitivity analysis alongside baseline forecasts is needed to guide policy design and communication. - Use the IPF and external tools more systematically: - FXI, CFMs, and MPMs may need more frequent consideration in both IT and non-IT frameworks. - Employ the IPF to select policy mixes and reduce financial frictions; improve data on cross-border flows and modalities for targeted FXI. - Study UMP design and exit in low-neutral-rate environments: - Appropriate design and calibration of UMP tools, including exit strategies and distributional implications, warrant further study. ### Quantitative and Research Priorities - Country-level granular analysis: - Heterogeneity in demographic pressures, climate vulnerabilities, exposure to fragmentation and AI readiness, fiscal space, and financial market depth will shape shock impacts and policy feasibility. - AI-related uncertainty: - Significant uncertainty about AI-driven growth; focus future research on mapping upside and downside risks, sectoral reallocation dynamics, and effects on fiscal and monetary policies and financial stability. - Model development: - Develop tractable ways to model cross-megatrend feedback loops and their implications for r−g differentials, natural rates, and term premia. - Practical toolkit improvements: - Empirical tools to distinguish shock nature in real time, improved data on cross-border flows (including digital and AI-enabled channels), and upgraded reserve adequacy frameworks. *Source: IMF Departmental Paper — “Resilience amid Structural Transformations,” Chapter 7: Policy Interactions and Trade-Offs.* ### Annex 1. Growth Impact of Structural ### Annex 1. Growth Impact of Structural Transformations: Methodology, Key Assumptions, and Main Results ### Overview: scope, horizon, and presentation - Growth impact measured as the difference between annual real GDP per capita growth including the trend and a “hypothetical baseline” without the trend. - Presents a “moderate” scenario (moderate-sized effects across possible scenarios). - Focuses on a 10-year horizon (2026–35); growth impact estimates are the average annual effect. - Sample covers 143 economies: 34 advanced economies (AEs), 68 emerging markets (EMs), and 41 low-income countries (LICs). --- ### Geoeconomic fragmentation — methodology and uncertainty - Methodology - Quantitative multicountry, multisector general equilibrium trade model following Caliendo and Parro (2015). - Calibrated to the 2023 EORA input–output matrix; covers 185 economies with 26 EORA sectors aggregated to 10. - Trade elasticities from Fontagné, Guimbard, and Orefice (2022) following Bolhuis, Chen, and Kett (2023). - Counterfactual GDP effects obtained by introducing changes in iceberg trade costs. (Footnote: iceberg trade costs encompass nontariff barriers and a degree of increased uncertainty.) - Moderate scenario assumptions - Prohibitively high trade costs between the US-EU and Russia, and in high-tech sectors between the US-EU and China. - The United States imposes a 10 percent increase in trade barriers on the rest of the world, including Chinese sectors other than high tech. - Trade barriers increase by 5 percent for all remaining trade. - High-tech sectors defined per Bolhuis, Chen, and Kett (2023) and Cerdeiro and others (2021) using OECD (2011) classification: electronics and machinery and transport equipment. - Uncertainty in estimates - Large variance in possible paths for trade barriers (example: 2025 US effective tariffs on imports from China rose from under 20 percent in January to over 100 percent in April, then dropped back to around 40–50 percent for the rest of the year). - Modeling choices matter: using iceberg trade costs vs. tariffs (tariff revenue rebated lump-sum) affects outcomes; Caliendo-Parro is a long-run static model that allows full sectoral adaptation and abstracts from dynamic capital accumulation. - Short-run trade elasticities may be smaller (Boehm, Levchenko, and Pandalai-Nayar 2023), implying larger transitional losses. - Other fragmentation channels not modeled here can be substantial: joint-channel aggregate losses up to 8.5 percent (Cerdeiro and others 2021); GDP losses from technological separation ranging about 0.4 percent to as much as 12 percent for most-exposed economies (Bekkers and Góes 2022); sustained FDI fragmentation could reduce global GDP by around 2 percent in the long run (IMF 2023b). - Trade policy uncertainty linked to weaker export growth and lower investment; effects of fragmentation in international payment systems remain an emerging research area. --- ### Demographics — methodology, moderate scenario, and uncertainties - Methodology - Based on a Cobb-Douglas production function with country-specific labor shares. - Effects calculated by accounting for changes in size and composition of labor force and population between t and t + 10 years. - K/L ratio assumed constant to isolate pure demographic effects (a ceteris paribus exercise). - Total population change = net births minus deaths + net migrations. Net migrations projected from historic trends and UN population division data (World Population Prospects Dataset 2024). - TFP affected by labor force composition: TFP reduced in countries with larger shares of older workers (coefficient denoted A). - Change in GDP between years t and t + i follows a linear approximation (see equation (1) in source); effect on growth rate is [y(t + i) / y(t) − 1] ÷ i. Change in GDP per capita follows equation (2) in source. - Moderate scenario - Uses UN central (median) cohort-component population projections (World Population Prospects 2024) by five-year age-groups to compute total and working-age population changes and applies equations (1) and (2). - Fertility projections use demographic-phase specific models (for most countries a Bayesian AR model); life expectancy based on expected healthcare and socioeconomic improvements; migration modeled with hierarchical Bayesian models. - Uncertainty in estimates - Over a 10-year horizon, main uncertainty arises from migration and, to a lesser extent, fertility; mortality stochasticity is small for closed countries. - Working-age population largely predetermined for horizons shorter than 15 years; uncertainty in GDP per capita growth due to demographics primarily from changes in number of dependents. - Deviations from the moderate scenario of about +0.2 and −0.2 percent per year for low- and high-fertility scenarios, respectively, while total GDP remains unchanged over the time period. - Average effects of demographic movements on GDP per capita over next 10 years by country group (table reproduced exactly) - Moderate | High Fertility | Low Fertility - AE −0.49 −0.69 −0.29 - EM − 0 .10 −0.33 0.14 - LIC 0.40 0.15 0.66 - Note: AE = advanced economy; EM = emerging market economy; LIC = low-income country. --- ### Artificial Intelligence (AI) — context, methodology, assumptions, and results - Context and motivation - Rapid advances in training and deploying large-language models (LLMs) and generative AI (GenAI) (ChatGPT launch 2022) raise prospects of a new wave of automation affecting labor productivity and growth. - Background of slowing labor productivity growth across AEs since mid-2000s; potential for a “J-curve” where initial investment and adjustment temporarily understate productivity gains. - Empirical exposure evidence and indices - Task-based exposure approaches identify tasks that are labor-displacing or labor-augmenting. - Felten, Raj, and Seamans (2023): AI Occupational Exposure (AIOE) index using O*NET and crowdsourced data—high exposure in highly paid, white-collar jobs. - Pizzinelli and others (2023): extend AIOE to account for complementary tasks requiring human supervision. - Eloundou and others (2023): GPT-4 and human annotations suggest around 80 percent of US workforce could have at least 10 percent of tasks affected; about 19 percent of workers may see at least 50 percent of tasks affected. - Methodology for cross-country GDP impact (extension of Acemoglu 2024) - Three variables required: (1) overall exposure to AI, (2) efficiency gains from AI (productivity gains and profitability), (3) speed of adoption (time horizon for gains). - Exposure to AI - Uses C-AIOE index from Cazzaniga and others (2024) at occupation level; rescales index to match mean and variance of exposure from Eloundou and others (2023) for the United States. - Aggregate exposure computed by weighting occupations by share in total wage bill using ILO data. - Exposure lower in most EMDEs because low-exposure occupations (e.g., agricultural) comprise a larger share of the wage bill. - Limitations: extrapolating US occupational task data may understate cross-country differences; sectoral composition assumed broadly constant over next decade. - Gains from AI: profitability and role of labor cost - Assumes 40 percent of tasks could be profitably automated in the next 10 years (assumption departs from Acemoglu (2024) and motivated by rapid expansion of AI offers and lower-cost models). - Cost of AI assumed similar across countries (abstracting from energy cost and regulation differences). - Labor costs vary: average wages in LICs are about 17 percent of the US level (PPP-adjusted), implying lower profitability of automation in LICs. - Occupation-level PPP-adjusted wage costs relative to the United States from the ILO used to create country-specific profitability estimates. - Productivity gains - Workplace experiments find performance improvements in the range of 14 percent (Brynjolfsson, Li, and Raymond 2023) to 55 percent (Peng and others 2023). - Robot automation cost savings estimated around 30 percent of labor costs (Acemoglu and Restrepo 2022). - Baseline productivity gain used: one-third over 10 years (higher than 27 percent used in Acemoglu 2024), following Aghion and Bunel (2024). - Same productivity gain assumed for all economies. - Adoption time - Uses IMF AI Preparedness Index (https://www.imf.org/external/datamapper/datasets/AIPI) to scale adoption lags relative to the United States. - Adoption time scaled such that the average lag is 3 years and the maximum lag is 10 years. - Rationale: somewhat faster diffusion than internet adoption estimates (Comin and Mestieri 2018: six years on average); exponential user growth suggests rapid diffusion, but lack of preparedness may slow adoption and deepen divergence. - Capital accumulation - Assumes constant capital-to-labor ratio in the medium term; more capital deepening would amplify GDP growth effects. - Results and cross-country heterogeneity - Most economies likely to see relatively modest effects from AI over next 10 years after accounting for exposure, labor cost differences, and adoption lags. - Median effects on real GDP per capita growth per year: - Median AE: 0.3 percentage points per year. - Median EM: 0.2 percentage points per year. - Median LIC: around 0.1 percentage points per year. - Upside risks: GenAI could substantially increase research productivity (e.g., drug discovery), and new products and services could improve welfare beyond measured productivity gains. - Downside risks: lower-than-expected productivity gains, sluggish adoption, adverse labor-market effects (rapid displacement, skill mismatch), cybercrime, higher energy costs. - Key assumptions and sensitive parameters (summary) - Share of tasks profitably automated in 10 years: 40 percent. - Baseline productivity gain over 10 years: one-third. - Adoption lag scaling: average 3 years; maximum 10 years. - LIC average wages about 17 percent of US level (PPP-adjusted). - Experimental evidence ranges for performance improvements: 14 percent to 55 percent; robots estimated cost savings ~30 percent of labor costs; Acemoglu (2024) used 27 percent. *Italic: Annex 1. Growth Impact of Structural Transformations: Methodology, Key Assumptions, and Main Results (source content).* ### Annex Figure 1.1. Cross-Country Distribution of Productivity Gains from AI over 10 Years—Effect of ### Annex Figure 1.1. Cross-Country Distribution of Productivity Gains from AI over 10 Years—Effect of Adjusting for Heterogeneity in Labor Cost and AI Preparedness ### Key empirical presentation - Graphic compares three series for increase in TFP over 10 years in percent: - Constant profitability - Adjusted for labor cost - Adjusted for AI preparedness - The figure explicitly shows results for the United States from Acemoglu (2024) for comparison. - Axis annotations as presented: "0", "0.2", "0.4", "0.6", "0.8", "1" and an additional sequence "00.511.522.5 3" (as shown in the source). - Source: IMF staff calculations. - Note: AI = artificial intelligence; TFP = total factor productivity. ### Interpretive notes on the AI estimates and heterogeneity adjustments - Adjustments for labor cost and AI preparedness alter cross-country distributions of projected TFP increases from AI over a 10-year horizon relative to a constant-profitability baseline. - United States outcome from Acemoglu (2024) is included as a reference point in the figure. ### Uncertainty and references - See Box 1 in the main text for further discussion of the literature and uncertainty in estimates. - Source line in figure: "IMF staff calculations." --- ### Climate change: methodology and uncertainty (contextual material in the same annex) ### Methodology for chronic and acute climate effects - Chronic GDP effects sourced from the IMF How-To Note (Centorrino, Massetti, and others 2025), based on Kahn and others (2021): panel econometric model exploiting deviations from historic temperature trends to identify GDP effects; projections based on forward-looking slow-moving temperature changes consistent with specific Shared Socioeconomic Pathways. - Main estimates represent GDP deviations relative to forward temperature projections corresponding to historic trends; adjustments applied to incorporate average GDP effect difference between the historic trend and no further temperature changes up to 2100. - Chronic damages capture persistent, long-term GDP losses from gradual changes in average temperature that affect productivity levels. - Acute damages reflect short-lived but potentially severe output losses from discrete extreme weather events; acute physical risk layer sourced from Phase IV NGFS output (NGFS portal referenced). Phase V output not used due to ongoing revisions. - NGFS acute data available for 49 economies (32 AEs and 17 EMs); when presenting total average effects, average chronic effects for full sample are combined with average acute effects for the subsample with acute data, implicitly setting missing acute observations to the income-group average. No acute effects available for LICs, so only chronic estimates are presented for LICs. ### Scenario definitions and quantitative uncertainty - Moderate chronic scenario: Shared Socioeconomic Pathway 2–4.5 ("current policies"), equating to 2.7 degrees of warming by 2100 relative to preindustrial period. - Moderate acute estimates: NGFS "Current Policies (Hot House World)" scenario with a temperature increase of 3°C by 2060. - Statistical uncertainty: parameter uncertainty captured by 95 percent confidence intervals around Kahn and others estimates; intervals widen over time and compound, often indistinguishable from zero in near term but becoming economically large by end of century. - Scenario and adaptation uncertainty dominates parameter uncertainty: projected GDP outcomes range from modest gains under strong mitigation to losses of several percentage points—reaching around 8 percent in high-warming, slow-adaptation scenarios—by 2100. - Example cross-study divergence: Bilal and Känzig (2024) estimate a permanent 1°C increase in global temperature reduces world GDP by roughly 20–30 percent in the long term. Survey evidence (Tol 2024; Morris and others 2025) shows estimates differ by an order of magnitude. --- ### Interactions between structural transformations (AI, GEF, demographics, climate) ### Qualitative interaction channels (summary from Annex Table 1.1) - Global Economic Fragmentation (GEF) × Climate: Low cooperation → higher emissions; reduced technology sharing; lower financing for climate adaptation in EMDEs. - GEF × Demographics: Immigration barriers reduce migration to aging economies; conflicts increase forced migration. - GEF × AI: Cross-border barriers slow AI development/adoption; reduced regulatory coordination; increased cyberattacks. - Climate × Demographics: Climate damages induce migration; population growth increases energy demand (and emissions); data center energy consumption (negative) and increased energy efficiency (positive) effects noted. - Demographics × AI: Older workers exposed to displacement (negative); AI may substitute for scarce younger workers (positive). ### Stylized quantitative interaction estimates (indicative) - Three specific interactions estimated for illustration: effect of GEF on demographics, effect of GEF on AI, and effect of AI and demographics on GEF. - GEF on demographics: shutting down all migration in the demographic model worsens outcomes where declining-population countries do not receive offsetting workers and growing-population countries retain excess workers assumed not to contribute to production. - GEF on AI: modeled as halving AI-related productivity growth relative to the moderate scenario (productivity growth = 50 percent of moderate scenario), worsening growth outcomes. - AI and demographics on GEF: modeled by doubling US import barriers from 10 percent to 20 percent and doubling barriers between third economies from 5 percent to 10 percent to capture increased trade barriers driven by AI competition and migration-related policy responses. ### Aggregate impact (moderate scenario plus interactions) - Stylized interactions add an additional negative effect of −0.41 percent on average annual per capita growth across economies to the moderate scenario effect (see Annex Figure 1.2). --- ### Annex Figure 1.2. Moderate Scenario plus Interactions, 2026−35 (Average yearly impact on GDP per capita growth) - Bar values shown in the figure: - Moderate: —0.05 - Interactions: —0.41 - Total: —0.46 - Y-axis labeled as "Average GDP impact (% of GDP)". - Sources for component estimates: demographics (staff calculations based on UN population projections with Cobb-Douglas production function); climate (chronic effects from Centorrino, Massetti, and others 2025; acute effects from NGFS); GEF (staff calculations using Caliendo and Parro (2015) long-run GE trade model); AI (staff calculations following Acemoglu (2024), incorporating labor market exposure from Cazzaniga and others (2024)); and IMF staff calculations. - Note: Sample covers 34 AEs, 68 EMs, and 41 LICs. Detailed methodologies described in Annex I. AEs = advanced economies; EMs = emerging markets; LICs = low-income countries. --- ### Economy coverage for GDP effect estimates - Total sample: 143 economies (starred economies are present in NGFS acute climate effects data set). - Sample counts by income group: 34 AEs, 68 EMs, 41 LICs. - Complete list of included economies (as presented): Australia (AE)*, Austria (AE)*, Belgium (AE)*, Canada (AE)*, Croatia (AE), Cyprus (AE), Czech Republic (AE)*, Denmark (AE)*, Estonia (AE)*, Finland (AE)*, France (AE)*, Germany (AE)*, Greece (AE)*, Hong Kong SAR (AE)*, Iceland (AE), Ireland (AE)*, Israel (AE), Italy (AE)*, Japan (AE)*, Korea (AE)*, Latvia (AE)*, Lithuania (AE)*, Luxembourg (AE), Netherlands (AE)*, Norway (AE)*, Portugal (AE)*, Singapore (AE)*, Slovak Republic (AE)*, Slovenia (AE)*, Spain (AE)*, Sweden (AE)*, Switzerland (AE)*, United Kingdom (AE)*, United States (AE)*, Algeria (EM), Angola (EM), Argentina (EM)*, Armenia (EM), Azerbaijan (EM), Belarus (EM), Belize (EM), Bolivia (EM), Bosnia and Herzegovina (EM), Botswana (EM), Brazil (EM)*, Brunei Darussalam (EM), Bulgaria (EM)*, Cabo Verde (EM), Chile (EM)*, China (EM)*, Colombia (EM), Costa Rica (EM), Dominican Republic (EM), Ecuador (EM), Egypt (EM)*, El Salvador (EM), Eswatini (EM), Fiji (EM), Georgia (EM), Guatemala (EM), Guyana (EM), Hungary (EM)*, India (EM)*, Indonesia (EM)*, Iran (EM), Iraq (EM), Jamaica (EM), Jordan (EM), Kazakhstan (EM), Lebanon (EM), Malaysia (EM)*, Mauritius (EM), Mexico (EM)*, Mongolia (EM), Montenegro (EM), Morocco (EM), Namibia (EM), North Macedonia (EM), Oman (EM), Pakistan (EM), Panama (EM), Paraguay (EM), Peru (EM), Philippines (EM), Poland (EM)*, Qatar (EM), Romania (EM)*, Russia (EM)*, Saudi Arabia (EM), Serbia (EM), South Africa (EM)*, Sri Lanka (EM), Suriname (EM), Thailand (EM), Trinidad and Tobago (EM), Tunisia (EM), Turkey (EM), Ukraine (EM), United Arab Emirates (EM), Uruguay (EM), Vanuatu (EM), Vietnam (EM)*, Afghanistan (LIC), Bangladesh (LIC), Benin (LIC), Bhutan (LIC), Burkina Faso (LIC), Burundi (LIC), Cambodia (LIC), Cameroon (LIC), Chad (LIC), Côte d’Ivoire (LIC), Djibouti (LIC), Ethiopia (LIC), Gambia (LIC), Ghana (LIC), Guinea (LIC), Honduras (LIC), Kenya (LIC), Kyrgyz Republic (LIC), Lao PDR (LIC), Lesotho (LIC), Liberia (LIC), Madagascar (LIC), Mali (LIC), Moldova (LIC), Mozambique (LIC), Myanmar (LIC), Nepal (LIC), Niger (LIC), Nigeria (LIC), Papua New Guinea (LIC), Rwanda (LIC), São Tomé and Príncipe (LIC), Senegal (LIC), Sierra Leone (LIC), Somalia (LIC), Tanzania (LIC), Togo (LIC), Uganda (LIC), Uzbekistan (LIC), Zambia (LIC), Zimbabwe (LIC). --- *Italic source attribution: IMF staff calculations; content drawn from the listed annex figures and accompanying annex text. ### Annex Table 2.1. Change in Total Public Expenditure in Percent of GDP, 80th Percentile within ### Annex Table 2.1. Change in Total Public Expenditure in Percent of GDP, 80th Percentile within ### Annex table values (as presented) - QuartileO n e Yea rFi ve Yea r - LICs11.63.4 - 22.84.5 - 34.05.2 - 49. 811.9 - EMs11.52.6 - 22.43 .1 - 33.65.8 - 45.95.9 ### Methodology to derive and apply caps - Public spending growth is constrained following the FfD approach, based on historical spending increases over 1999–2019. - Common caps for EMs and LICs are derived using the 80th percentile of positive one-year spending changes (2009–19), aggregated through quartiles to obtain group-level one-year and five-year caps. Final caps by group are reported in Annex Table 2.1. - The one-year cap applies in the first year, whereas spending in years 2–4 can increase cumulatively up to the five-year cap. - From year 6, a new one-year increase is allowed, and a new five-year cap, incremental to the previous, is applied from year 7. This pattern repeats every five years until 2044. - The spending caps are applied to the values of the ratio of spending to GDP given by Carapella and others (2023), which are calibrated to be constant as a percent of GDP annually. - For 2025, the additional spending ratio is capped at the minimum of the unconstrained target and a one-year ceiling. From 2026 onward, each year’s constrained ratio is the minimum of: - (1) the unconstrained target for that year, - (2) a one-year increment cap equal to last year’s constrained level plus one-year ceiling, and - (3) a cumulative cap tied to five-year blocks multiples since 2025. ### Fiscal multipliers and GDP effects - Fiscal multipliers applied: 0.4 for first year, 0.5 for second, 0.4 for third, 0.3 for fourth, 0.3 for fifth, and 0.1 for sixth year. - Multipliers are applied to total output with a contemporaneous effect as well as lagged effects up to six years. The cumulative effect is computed by applying the relevant lagged spending amounts to each year’s multiplier and summing across the six-year window. - The cumulative effect is then added to the exogenous GDP to generate an endogenous GDP path that reflects growth-enhancing effects of SDG-related spending. - The approach assumes these multipliers applied to total spending on total GDP (rather than to spending-to-GDP ratios), producing an endogenous GDP path higher than in the Finance for Development paper because of a longer time horizon. - Alternative recursive implementation (updating GDP year by year to allow feedback and dynamic compounding) was considered but not used to avoid overestimation; the chosen cumulative implementation aims for simplicity. ### Spending composition and proportionality assumptions - Nominal spending on defense, health, pensions, and climate is assumed to scale proportionally with endogenous GDP, leaving spending-to-GDP ratios unchanged. - Given the smoothing profile and extended implementation horizon, average spending ratios are intentionally lower and not directly comparable to Carapella and others (2023) and the FfD for both LICs and EMs. - Estimates are indicative and should not be used for country-level policy analysis. For the MSD estimates, the average per country is taken across the 10- and 20-year time horizons. All estimates are expressed as an annual percent of GDP relative to 2024. ### Smoothing and ramp-up assumptions - To smooth the spending profile over 2025–44 and avoid front- or back-loaded paths, the SDG total spending (in US$) for countries with FfD data is distributed over 2025–44 to create a smooth spending ramp-up reflecting implementation constraints and institutional capacity. - A Gompertz function is used to smooth the distribution: w(t) = a ⋅ exp[−b ⋅ e{−ct}]. - Parameters for LICs and EMs are calibrated through iterative testing to generate a realistic and internally consistent spending ramp-up in the absence of observed year-by-year execution patterns. - For countries without FfD data and without a smoothing profile in Carapella and others (2023), constrained annual constant estimates are used. ### Key implementation notes and caveats - The framework incorporates the first-order effect of fiscal multipliers by allowing GDP to grow above baseline in response to higher public spending while keeping nominal spending constant. This reduces spending-to-GDP ratios over time and reflects potential easing of relative pressures because of higher growth. - The framework does not endogenize second-round feedback such as higher nominal spending from income-linked expenditures, increased financing needs, or interest rate responses. Rising debt and debt service costs—particularly in a high-interest rate environment—could become new sources of fiscal pressure; these dynamics are not fully captured. Results are indicative and partial equilibrium in nature. - Gaspar and others (2019) indicate that even for the case of human capital spending, spending may have to rise before 2030, making the ramp-up assumption consistent with that evidence. *Source: Finance for Development paper, Annex I. Box 2. Figure 1. The 4th Financing for Development Conference—Contribution of the IMF to the International Financing for Development Agenda—Annexes in: Policy Papers Volume 2025 Issue 022 (2025).* ### Annex Table 3.1. Descriptive Statistics in the MSD Simulations ### Annex Table 3.1. Descriptive Statistics in the MSD Simulations ### Descriptive statistics (tax capacity, optimal tax, and tax gap) - Table header mapping (as presented): Tax Capacity — Optimal Tax — Tax Gap — Percent of Tax Gap Allowed to Vary in MSD Simulations without Human Capital Investment — Percent of Tax Gap Allowed to Vary in MSD Simulations with Human Capital Investment. Columns reported as: Mean, Median, Mean, Median, Mean, Median, Mean, Median, Mean, Median. - AE (advanced economy): - Tax Capacity Mean: 51.5% - Tax Capacity Median: 48.3% - Optimal Tax Mean: 47.3% - Optimal Tax Median: 47.1% - Tax Gap Mean: 4.2% - Tax Gap Median: −0.7% - Percent of Tax Gap Allowed to Vary without Human Capital Investment Mean: 4.2% - Percent of Tax Gap Allowed to Vary without Human Capital Investment Median: −0.7% - Percent of Tax Gap Allowed to Vary with Human Capital Investment Mean: 4.2% - Percent of Tax Gap Allowed to Vary with Human Capital Investment Median: −0.7% - EM (emerging market): - Tax Capacity Mean: 39.2% - Tax Capacity Median: 35.4% - Optimal Tax Mean: 31.5% - Optimal Tax Median: 30.5% - Tax Gap Mean: 7.7% - Tax Gap Median: 5.2% - Percent of Tax Gap Allowed to Vary without Human Capital Investment Mean: 3.9% - Percent of Tax Gap Allowed to Vary without Human Capital Investment Median: 2.6% - Percent of Tax Gap Allowed to Vary with Human Capital Investment Mean: 5.0% - Percent of Tax Gap Allowed to Vary with Human Capital Investment Median: 3.4% - LIC (low-income country): - Tax Capacity Mean: 35.5% - Tax Capacity Median: 31.6% - Optimal Tax Mean: 24.7% - Optimal Tax Median: 21.6% - Tax Gap Mean: 10.9% - Tax Gap Median: 8.1% - Percent of Tax Gap Allowed to Vary without Human Capital Investment Mean: 2.6% - Percent of Tax Gap Allowed to Vary without Human Capital Investment Median: 2.0% - Percent of Tax Gap Allowed to Vary with Human Capital Investment Mean: 5.3% - Percent of Tax Gap Allowed to Vary with Human Capital Investment Median: 4.0% - Notes and definitions (as provided): - The table shows the mean and median for maximum tax capacity and optimal tax as a percent of GDP in the model, per income group. - The first and second columns show the mean and median value of taxation capacity as a percent of GDP by income group. - The third and fourth columns show the mean and median of optimal taxation, which is backed out from the model (see equation (1) in Annex III) by income group. - The last four columns show the mean and median of the additional tax as a percent of GDP that is allowed to be raised on top of the baseline optimal taxation per income group. A negative gap means the tax capacity is binding. - AE = advanced economy; EM = emerging market; LIC = low-income country; MSD = maximum sustainable debt. ### Effects of structural transformations on growth and the natural rate (Annex Table 3.2) - Sample note (as provided): These data are for the sample of 131 economies used in the MSD simulations. These estimates are added to the baseline parameters of the model. Footnote: This table summarizes estimates used in the MSD analysis for 128 countries so the spending estimates may differ from the ones reported in text for the full sample of 190 countries. - Growth Effects (Average Annual Percentage Point Change) — columns: Demographics, GEF, Climate, AI, All - AE: −0.33, −0.20, −0.02, 0.31, −0.24 - EM: 0.46, −0.18, −0.02, 0.20, 0.46 - LIC: 2.12, −0.20, −0.02, 0.13, 2.02 - Natural Rate Effects (Percentage Point Change) — columns: Demographics, Climate, GEF, AI, All - AE: −0.02, −0.08, 0.17, 0.15, 0.22 - EM: −0.5, −0.08, 0.02, 0.1, −0.46 - LIC: −0.5, −0.08, 0.03, 0.1, −0.45 - Additional notes: - Defense assumes for NATO countries a gradual target of 3.5 percent, and 2 percent for all other economies. Climate estimates include only chronic effects. AE = advanced economy; EM = emerging market; GEF = global economic fragmentation; LIC = low-income country; MSD = maximum sustainable debt; NATO = North Atlantic Treaty Organization. - The sample references in different places: 131 economies for MSD simulations; a footnote clarifies 128 countries for the MSD analysis estimates; elsewhere the sample includes 129 economies for figure estimations. ### Key qualitative findings on fiscal space and MSD - Spending pressures from transformations risk reducing fiscal space unless accompanied by the necessary tax adjustments. - For the climate megatrend, the adaptation spending assumed in the fiscal scenario is assumed to offset acute damages. - Median MSD estimates and accounting for additional human capital investments for the moderate scenario assumes higher tax increases if the demographic dividend is to be specified. - Sensitivity analysis on more ambitious infrastructure and human capital investments, under the current assumptions of tax increases, will likely reduce fiscal space and may necessitate further revenue mobilization. - Increasing primary balance is key to cushioning the MSD: sensitivity tests assuming no tax increases are allowed show a deterioration for the MSD for all income groups. ### Annex Figures (summary of content and methodology) - Annex Figure 3.3: Effect of Structural Transformations on MSD Levels across Economies, by Income Group (panels for Moderate Scenario, trimming outliers, Moderate Scenario with Human Capital, Sensitivity with Higher Infrastructure and Human Capital). - Change in MSD measured in percentage points across AE, EM, LIC; contributions estimated through a Shapley value decomposition. Panel 4 shows sensitivity when higher estimates on infrastructure and human capital are assumed without further tax increases. Climate estimates include only chronic effects. A higher increase in taxation is allowed for human capital when added (see Annex Table 3.1). - Sample includes 129 economies (as noted in figure caption). - Annex Figure 3.4: Contributions of Structural Transformations to Shifts in MSD Levels — chart shows median and 25th/75th percentiles with contributions winsorized at 5th and 95th percentiles and estimated based on a Shapley value decomposition. Climate estimates include only chronic effects. *Source: IMF staff calculations.* ### Annex 4. Fiscal Frameworks: Additional Spending and Debt Dynamics ### Model overview and key mechanisms - Model used: developed by Nguyen, Lam, and Sher (2025); integrates literature on public investment and output multipliers, endogenous relationship between government debt and borrowing costs, nonlinear/time-varying elasticity of interest rates to debt levels, and pass-through of sovereign borrowing costs to private sector funding costs. - Core mechanisms summarized: - Endogeneity of sovereign borrowing costs: household demand for safe assets and diminishing convenience benefits as debt rises (building on Mian, Straub, and Sufi (2025)). - Time-varying elasticity φt linked to probability of debt distress ρt, where ρt is logistic and lies between 0 and 1; in case of debt distress ρt = 1, premium assumed to go up by 1 percentage point for each 10 percentage points increase in debt (Δ = 0.1). Specification similar to Bi and Traum (2012). - Spillover to private sector cost of funds following Burriel and others (2020) and empirical sensitivity of corporate yields to sovereign yields about 0.5 (Bevilaqua, Hale, and Tallman (2020)). - Key equations referenced (symbols preserved as in source): Euler equation (4); convenience yield link (5); time-varying elasticity relation (6); expression for φt (7); return on investment with wedge ψ (8); wedge definition ψφ (9). ### Calibration (key parameters distinguishing country groups) - Table of parameters (AE, EM, LIC): - Prob. of debt distress: intercept parameter: −5.2, −4.3, −4.4 - Prob. of debt distress: slope parameter: 3.5, 3.7, 5.5 - Consumption tax rate (SS): 20%, 15%, 1.5% - Labor income tax rate (SS): 25%, 20%, 0.7% - Government consumption as percent of GDP (SS): 20%, 15%, 10% - Government investment as percent of GDP (SS): 4%, 6.5%, 15% - Steady-state growth: 2.1%, 3.7%, 4.8% - Note: AE = advanced economy; EM = emerging market; LIC = low-income country. SS = steady state. ### Simulation assumptions - Additional expenditure decomposed into two components: government consumption (including health, pension, and half of defense spending) and government investment (including infrastructure, climate adaptation, and half of defense spending). - Potential growth assumptions over 10 years: - Advanced economy: a gradual decrease in potential growth, resulting in a drop of 0.32 percent by year 10. - Emerging market: a small increase in potential growth by 0.46 percent by year 10. - Low-income country: an increase in potential growth of 2.05 percent by year 10. ### Implications from simulations (summary) - For a “nonvulnerable” (advanced-economy–type) calibration facing a large adverse cost-push shock: - Stronger monetary tightening alone stabilizes inflation but raises policy rates substantially, increasing government debt and reducing private consumption. - Faster fiscal adjustment via discretionary cuts to government consumption can support monetary policy, allowing lower policy-rate increases, similar inflation/output outcomes, improved private consumption, and lower debt relative to the initial steady state, although it can depreciate the exchange rate slightly and offset some inflation benefits. *Source: IMF staff calculations.* ### Annex Figure 6.1. Impulse Response Functions for Scenario A—Nonvulnerable Economy ### Annex Figure 6.1. Impulse Response Functions for Scenario A—Nonvulnerable Economy ### Scenario A — Nonvulnerable economy (figure overview) - Three policy variants illustrated: Slow fiscal adjustment; Faster fiscal adjustment; Slow fiscal adjustment and more aggressive monetary policy. - Note: Response to large 3 standard deviation shock at time t = 1. All variables are reported as deviations from baseline paths. - Variables shown (labels preserved): Output; Core inflation (APR); Transfers (% of trend GDP); Government debt (% of trend GDP); Net foreign assets (% of trend GDP); Tax revenue (% of trend GDP); Gov. consumption (% of trend GDP); UIP premium (APR); Real exchange rate; Policy rate (APR); Foreign output; Foreign core inflation (APR); Consumption; Real net exports (% of trend GDP); Nominal trade balance (% of trend GDP). - Acronyms: APR=annual percentage rate; UIP=uncovered interest parity; NFA=net foreign assets. ### Scenario B — Vulnerable economy (key characteristics and implications) - Replacement of the nonvulnerable “home economy” with a vulnerable economy featuring: - Shallow FX markets (a larger FX market frictions parameter). - Significant foreign debt: negative NFA of 95 percent of GDP. - Implications described: - The real exchange rate moves less under fiscal consolidation because of offsetting effects of interest rate differentials and the improvement in the NFA position. - Fiscal restraint is even more effective in curbing inflationary pressures than in the small open advanced economy case. - Note: Response to large 3 standard deviation shock at time t = 1. All variables are reported as deviations from baseline paths. ### Scenario C — Repeated supply shocks (dynamics and policy trade-offs) - Setup: - Initial global cost‑push shock at t = 1, followed by a second identical cost‑push shock at varying subsequent times. - Shock size: moderate 2 standard deviation shock at time t = 1, followed by an identical subsequent shock in varying periods. - Key findings: - When the second shock occurs four quarters after the first (blue line), the initial inflation response is just past its peak. - The secondary inflation response is broadly similar to the first but requires a much larger policy rate response: roughly 4 percentage points versus just over 2. - Larger policy‑rate responses have adverse consequences for output and private consumption. - The initial shock elevates inflation and brings the economy closer to the nonlinear section of the Phillips curve and elevated wage indexation, increasing the risk that the second shock creates much larger and more persistent inflationary pressure if left unchecked. - Policy implication: - Results support monetary policy being more wary about looking through temporary supply shocks when risks of repeated shocks increase. - Note: All variables are reported as deviations from baseline paths. APR=annual percentage rate; UIP=uncovered interest parity; NFA=net foreign assets. ### Scenario D — Weak versus strong monetary–fiscal frameworks - Regime comparison: - Low‑debt regime: public debt and net foreign liabilities are low; monetary and fiscal frameworks are strong and credible; inflation expectations well anchored. - High‑debt regime: public and external debts elevated; policy frameworks weaker; inflation expectations less firmly anchored, allowing greater intrinsic inflation persistence. - Common assumption: - FX markets are assumed to be deep in both regimes (deviations from UIP remain limited; only small movements in UIP premia). - No endogenous term premia introduced; all other structural and policy assumptions held fixed across regimes. - Key finding: - Even with deep FX markets, weaker monetary–fiscal frameworks in the high‑debt economy amplify inflation persistence and worsen policy trade‑offs relative to the low‑debt, strong‑framework case. - Note: Response to large 3 standard deviation shock at time t = 1. All variables are reported as deviations from baseline paths. APR=annual percentage rate; UIP=uncovered interest parity; NFA=net foreign assets. ### Scenario E — Using a discretionary spending cut to control debt - Policy constraint: - Debt constrained to return to its preshock level within five years. - Policy variants illustrated: - Red line: moderate discretionary cut to government consumption (Disc. G cut). - Outcome: worsens output but somewhat improves inflation outcomes. - Yellow line: mix of tighter monetary policy stance (higher real rates) and even tighter fiscal stance to keep debt unchanged (Disc. Tighter MF). - Outcome: succeeds in lowering inflation but at a more significant output cost. - Trade‑off and contextual considerations: - Trade‑offs are palpable; appropriate policy response depends on how limited fiscal space is and how close the economy is to inflation expectations de‑anchoring. - Note: Response to large 3 standard deviation shock at time t = 1. All variables are reported as deviations from baseline paths. APR=annual percentage rate; UIP=uncovered interest parity; NFA=net foreign assets. *Source: IMF staff calculations.* ### Executive Summary. Ottawa: Bank of Canada. ### Executive Summary. Ottawa: Bank of Canada. ### References cited (selection) - Bank of England. 2024. Forecasting for Monetary Policy Making and Communication at the Bank of England: A Review. London: Bank of England. - Barro, R. J. 1979. “On the Determination of the Public Debt.” Journal of Political Economy 87 (5): 940–71. - Beaudry, P., T. J. Carter, and A. Lahiri. 2023. “The Central Bank’s Dilemma: Look Through Supply Shocks or Control Inflation Expectations?” NBER Working Paper 31741, National Bureau of Economic Research, Cambridge, MA. - Bekkers, E., and C. Góes. 2022. “The Impact of Geopolitical Conflicts on Trade, Growth, and Innovation: An Illustrative Simulation Study.” VoxEU (CEPR), March 29. - Benitez, J. C., M. Mansour, M. Pecho, and C. Vellutini. 2023. “Building Tax Capacity in Developing Countries.” IMF Staff Discussion Note 2023/006, International Monetary Fund, Washington, DC. - Bevilaqua, J., G. Hale, and E. Tallman. 2020. “Corporate Yields and Sovereign Yields.” Journal of International Economics 124 (6): 103304. - Beyer, R. C. M., and L. Brandão-Marques. 2025. “Real Equilibrium Interest Rates in the Euro Area.” IMF Working Paper 2025/123, International Monetary Fund, Washington, DC. - Bi, H., and N. Traum. 2012. “Estimating Sovereign Default Risk.” American Economic Review 102 (3): 161–66. - Bianchi, F., H. Kung, and M. Tirskikh. 2023. “The Origins and Effects of Macroeconomic Uncertainty.” Quantitative Economics 14:855 –96. - Bilal, A., and D. R. Känzig. 2024. “The Macroeconomic Impact of Climate Change: Global vs. Local Temperature.” Quarterly Journal of Economics 139 (3): 1887–929. - Blanchard, O. 2005. “Fiscal Dominance and Inflation Targeting: Lessons from Brazil.” In Inflation Targeting, Debt, and the Brazilian Experience, edited by F. Giavazzi, I. Goldfajn, and S. Herrera. Cambridge, MA: MIT Press. - Blanchard, O. J., and B. S. Bernanke. 2023. “What Caused the US Pandemic-Era Inflation.” NBER Working Paper 31417, National Bureau of Economic Research, Cambridge, MA. - Boehm, C. E., A. A. Levchenko, and N. Pandalai-Nayar. 2023. “The Long and Short (Run) of Trade Elasticities.” American Economic Review 113 (4): 861–905. - Bogmans, C., P. Gomez-Gonzalez, G. Ganpurev, G. Medina, A. Pescatori, and S. D. Thube. 2025. “Power Hungry: How AI Will Drive Energy Demand.” IMF Working Paper 2025/081, International Monetary Fund, Washington, DC. - Bolhuis, M., J. Chen, and B. Kett. 2023. “Fragmentation in Global Trade: Accounting for Commodities.” IMF Working Paper 2023/073, International Monetary Fund, Washington, DC. - Bolhuis, M. A., C. M. Rui, and H. Thorell. Forthcoming. “Consequences of Undermining Central Bank Independence: Evidence from Governor Transitions.” IMF Working Paper, International Monetary Fund, Washington, DC. - Brand, C., N. Lisack, and F. Mazelis. 2024. “Estimates of the Natural Interest Rate for the Euro Area: An Update.” ECB Economic Bulletin Issue 1/2024, European Central Bank, Frankfurt. - Brandão-Marques, L., M. Casiraghi, G. Gelos, O. Harrison, and G. Kamber. 2023. “Is High Debt Constraining Monetary Policy? Evidence from Inflation Expectations.” IMF Working Paper 2023/33, International Monetary Fund, Washington, DC. - Brandão-Marques, L., M. Casiraghi, G. Gelos, G. Kamber, and R. Meeks. 2024. “Negative Interest Rate Policies: A Survey.” Annual Review of Economics 16:305–28. - Brandão-Marques, M. L., M. R. Meeks, and V. Nguyen. 2024. “Monetary Policy with Uncertain Inflation Persistence.” IMF Working Paper 2024/047, International Monetary Fund, Washington, DC. - Briggs, J., and D. Kodnani. 2023. “The Potentially Large Effects of Artificial Intelligence on Economic Growth.” Goldman Sachs Global Investment Research, March 26. - Brollo, F., E. Dabla-Norris, R. de Mooij, D. Garcia-Macia, T. Hanappi, L. Liu, and A. D. M. Nguyen. 2024. “Broadening the Gains from Generative AI: The Role of Fiscal Policies.” IMF Staff Discussion Note 2024/002, International Monetary Fund, Washington, DC. - Brynjolfsson, E., D. Li, and L. Raymond. 2023. “Generative AI at Work.” NBER Working Paper 31161, National Bureau of Economic Research, Cambridge, MA. - Brynjolfsson, E., D. Rock, and C. Syverson. 2021. “The Productivity J-Curve: How Intangibles Complement General Purpose Technologies.” American Economic Journal: Macroeconomics 13 (1): 333–72. - Budina, N., C. H. Ebeke, F. Jaumotte, A. Medici, A. J. Panton, M. M. Tavares, and B. Yao. 2023. “Structural Reforms to Accelerate Growth, Ease Policy Trade-offs, and Support the Green Transition in Emerging Market and Developing Economies.” IMF Staff Discussion Note 2023/007, International Monetary Fund, Washington, DC. - Burke, M., S. Hsiang, and E. Miguel. 2015. “Global Non-linear Effect of Temperature on Economic Production.” Nature 527 (7577): 235–39. - Burriel, P., C. Checherita-Westphal, P. Jacquinot, M. Schön, and N. Stähler. 2020. “Economic Consequences of High Public Debt: Evidence from Three Large Scale DSGE Models.” ECB Working Paper 2450, European Central Bank, Frankfurt. - Cacciatore, M., B. Feunou, and G. K. Ozhan. 2024. “The Neutral Interest Rate: Past, Present and Future.” Bank of Canada Staff Discussion Paper 2024-3, Bank of Canada, Ottawa. - Cairó, I., C. Gust, J. Hebden, E. Herbst, S. Konzem, and G. Nicolò. 2025. “Risk-Adjusted Optimal Policy for Scenario Analysis.” IMF Working Paper 2025/71, International Monetary Fund, Washington, DC. - Caliendo, L., and F. Parro. 2015. “Estimates of the Trade and Welfare Effects of NAFTA.” Review of Economic Studies 82 (1): 1–44. - Canova, F., and G. De Nicoló. 2002. “Monetary Disturbances Matter for Business Fluctuations in the G-7.” Journal of Monetary Economics 49 (6): 1131–59. - Cao, Y., R. Lam, W. Jiang, and N. Wang. 2025. “Debt Carrying Capacity Across Countries: An Assessment Using p-Theory.” IMF Working Paper, International Monetary Fund, Washington, DC. - Carapella, P., T. Mogues, J. Pico-Mejia, and M. Soto. 2023. “How to Assess Spending Needs of the Sustainable Development Goals: The Third Edition of the IMF SDG Costing Tool.” IMF How-To Note 2023, International Monetary Fund, Washington, DC. - Cardozo, P., A. Fernández, J. Jiang, and F. D Rojas. 2024. “On Cross-Border Crypto Flows: Measurement Drivers and Policy Implications.” IMF Working Paper 2024/261, International Monetary Fund, Washington, DC. - Carter, T. J., G. Kamber, and J. Otten. 2026. “How to Design and Review Inflation Targets in Emerging Market and Developing Economies.” IMF How-To Note 2026/004, International Monetary Fund, Washington, DC. - Casiraghi, M., K. Habermeier, and T. Harjes. 2022. Choice of Exchange Rate Arrangement: Monetary and Capital Markets Department Technical Assistance Handbook. Washington, DC: International Monetary Fund. - Cazzaniga, M., F. Jaumotte, L. Li, G. Melina, A. J. Panton, C. Pizzinelli, E. Rockall, and M. M. Tavares. 2024. “Gen-AI: Artificial Intelligence and the Future of Work.” IMF Staff Discussion Note 2024/001, International Monetary Fund, Washington, DC. - Centorrino, S., A. Diakantoni, A. Keck, M. Ruta, M. Sztajerowska, and Y. Wei. 2025. “Measuring Global Trade Policy Activity.” IMF Working Paper 2025/220, International Monetary Fund, Washington, DC. - Centorrino, S., E. Massetti, M. Raissi, and F. Tagklis. 2025. “How to Include the Effects of Rising Temperatures in Long-Term GDP Projections.” IMF How-To Note 2025/009, International Monetary Fund, Washington, DC. - Cerdeiro, D. A., R. Mano, J. Eugster, M. D. V. Muir, and M. S. J. Peiris. 2021. Sizing Up the Effects of Technological Decoupling. Washington, DC: International Monetary Fund. - Cerutti, E. M., A. I. Garcia Pascual, Y. Kido, L. Li, G. Melina, M. Mendes Tavares, and P. Wingender. 2025. “Mind the Gap.” IMF Working Paper 2025/076, International Monetary Fund, Washington, DC. - Chen, J., E. Dabla-Norris, C. Goncalves, Z. Jakab, and J. Lindé. 2023. “Can Fiscal Consolidation Help Central Banks’ Fight Inflation?” IMF Working Paper 2023/260, International Monetary Fund, Washington, DC. - Christensen, J. H. E., and G. D. Rudebusch. 2012. “The Response of Interest Rates to U.S. and U.K. Quantitative Easing.” Working Paper 2012-06, Federal Reserve Bank of San Francisco. - Coibion, O., Y. Gorodnichenko, and M. Weber. 2021. “Fiscal Policy and Households’ Inflation Expectations: Evidence from a Randomized Control Trial.” NBER Working Paper 28485, National Bureau of Economic Research, Cambridge, MA. - Comin, D., and M. Martí. 2018. “If Technology Has Arrived Everywhere, Why Has Income Diverged?” American Economic Journal: Macroeconomics 10 (3): 137–78. - Corsetti, G., K. Kuester, A. Meier, and G. J. Müller. 2013. “Sovereign Risk, Fiscal Policy, and Macroeconomic Stability.” Economic Journal 123 (566): F99–F132. - De Long, J. B., A. Shleifer, L. H. Summers, and R. J. Waldmann. 1990. “Noise Trader Risk in Financial Markets.” Journal of Political Economy 98 (4): 703–38. - Dietz, S, and N. Stern. 2015. “Endogenous growth, convexity of damage and climate risk: How Nordhaus’ framework supports deep cuts in carbon emissions.” The Economic Journal 125(583): 574–620. - Eble, S., A. Pitt, I. Bunda, O. E. Adilbish, N. Budina, G. H. Hong, M. Malak, S. Mohona, A. Myrvoda, and K. Primus. 2025. “Long-Term Spending Pressures in Europe.” IMF Departmental Paper 2025/002, International Monetary Fund, Washington, DC. - Eloundou, T., S. Manning, P. Mishkin, and D. Rock. 2023. “GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models.” Technical Report, arXiv 2023. - Emeksiz, E. O., G. Kamber, J. Otten, and G. K. Pasricha. 2026. “Still Packing a Punch: Monetary Policy Transmission in a New Cross-Country High-Frequency Dataset.” IMF Working Paper 2026-018, International Monetary Fund, Washington, DC. - Erceg, C. J., J. Lindé, and M. Trabandt. 2024. “Monetary Policy and Inflation Scares.” IMF Working Paper 2024/260, International Monetary Fund, Washington, DC. - European Central Bank (ECB). 2024. “Artificial Intelligence: A Central Bank’s View.” Keynote Speech by Piero Cipollone, Member of the Executive Board of the European Central Bank, at the National Conference of Statistics on Official Statistics at the Time of Artificial Intelligence, Rome, July 4. - Eyraud, L., A. Baum, A. Hodge, M. Jarmuzek, H. E. Ture, S. Mbaye, and Y.-B. Kim. 2018. “How to Calibrate Fiscal Rules: A Primer.” IMF How-To Note 2018/002, International Monetary Fund, Washington, DC. - Felten, E. W., M. Raj, and R. Seamans. 2023. Occupational Heterogeneity in Exposure to Generative AI. SSRN. - Fernald, J., R. Inklaar, and D. Ruzic. 2023. “The Productivity Slowdown in Advanced Economies: Common Shocks or Common Trends?” Working Paper 2023-07, Federal Reserve Bank of San Francisco. - Feyrer, J. 2007. “Demographics and Productivity.” Review of Economics and Statistics 89 (1): 100 –109. - Filippucci, F., P. Gal, and M. Schief. 2024. “Miracle or Myth? Assessing the Macroeconomic Productivity Gains from Artificial Intelligence.” OECD Artificial Intelligence Paper 29, OECD Publishing, Paris. - Flores, E., P. Gupta, Y. Lu, P. A. Medas, D. Prihardini, H. Selim, W. Xin, and M. Yabara. 2024. “Upgrading Fiscal Frameworks in Asia-Pacific.” IMF Departmental Paper 2024-008, International Monetary Fund, Washington, DC. - Fontagné, L., H. Guimbard, and G. Orefice. 2022. “Tariff-Based Product-Level Trade Elasticities.” Journal of International Economics 137:10 3 593 . - Freedman, C., and M. D. Laxton. 2009. Inflation Targeting Pillars-Transparency and Accountability. Washington, DC: International Monetary Fund. - Furceri, D., D. Giannone, F. Kisat, W. Raphael Lam, and H. Li. 2025. “Debt-at-Risk.” IMF Working Paper 2025-086, International Monetary Fund, Washington, DC. - Gaspar, V., M. Garcia-Escribano, D. Prady, and M. Soto. 2019. “Fiscal Policy and Development: Human, Social, and Physical Investments for the SDGs.” Staff Discussion Notes 2019, International Monetary Fund, Washington, DC. - Gaspar, V., M. Mansour, and C. Vellutini. 2023. “Countries Can Tap Tax Potential to Finance Development Goals.” IMF Blog, September 19. - Gnan, E., and R. Holzmann. 2023. “Should Central Banks Raise Inflation Targets?” SUERF Policy Note Issue 304, March. - Gopinath, G. 2022. “How the Pandemic and War Shape Future Monetary Policy.” Jackson Hole Symposium: Federal Reserve Bank of Kansas City. - Gopinath, G. 2024. “Crisis Amplifier? How to Prevent AI from Worsening the Next Economic Downturn.” Speech at AI for Good Global Summit, Geneva, May 30. - Gornicka, L., S. R. Mirza, V. Nguyen, and J. Vandenbussche. Forthcoming. “When Policy Bites: State-Dependent Monetary Policy Transmission in Emerging Markets.” IMF Working Paper, International Monetary Fund, Washington, DC. - Gourinchas, P. O. 2026. “Adequate Reserves Shield Economies from Shocks and Strengthen Resilience.” IMF Blog, March 5. - Guerrieri, V., G. Lorenzoni, and I. Werning. 2025. “Global Price Shocks and International Monetary Coordination.” NBER Working Paper 33840, National Bureau of Economic Research, Cambridge, MA. - Harding, M., J. Lindé, and M. Trabandt. 2023. “Understanding Post-COVID Inflation Dynamics.” Journal of Monetary Economics 140:S101–S118. - Hartmann, P., and V. Maver. 2025. “Implications of Artificial Intelligence for Monetary Policy: A First Conceptual Assessment.” SUERF Policy Brief 1080. - Hiraki, K., and M. Rhouzlane. Forthcoming. “How to Operationalize an Exchange Rate Peg.” IMF How-To Note, International Monetary Fund, Washington, DC. - Hofman, D., and G. Kamber. 2020. “Unconventional Monetary Policy in Emerging Market and Developing Economies.” IMF MCM Special Series on COVID-19, International Monetary Fund, Washington, DC. - Howard, P. H., and T. Sterner. 2017. “Few and Not So Far Between: A Meta-Analysis of Climate Damage Estimates.” Environmental and Resource Economics 68 (1): 197–225. - Howard, P. H., and D. Sylvan. 2020. “Wisdom of the Experts: Using Survey Responses to Address Positive and Normative Uncertainties in Climate-Economic Models.” Climatic Change 162 (2): 213–32. - Imam, P. 2013. “Shock from Graying: Is the Demographic Shift Weakening Monetary Policy Effectiveness.” IMF Working Paper 2013/191, International Monetary Fund, Washington, DC. - International Monetary Fund (IMF). 2011. “Assessing Reserve Adequacy.” IMF Policy Paper, Washington, DC, February. https://www.imf.org/external/np/pp/eng/2013/111313e.pdf - International Monetary Fund (IMF). 2013. “Assessing Reserve Adequacy—Further Considerations: Supplementary Information.” IMF Policy Paper, Washington, DC, November 15. https://www.imf.org/external/np/pp/eng/2014/121914.pdf - International Monetary Fund (IMF). 2014. “Assessing Reserve Adequacy—Specific Proposals.” IMF Policy Paper, Washington, DC, December 19. https://www.imf.org/external/np/pp/eng/2016/060316.pdf - International Monetary Fund (IMF). 2015. “Evolving Monetary Policy Frameworks in Low-Income and Other Developing Countries.” IMF Policy Papers, Washington, DC. - International Monetary Fund (IMF). 2016. “Guidance Note on the Assessment of Reserve Adequacy and Related Considerations.” IMF Policy Paper, Washington, DC, June. - International Monetary Fund (IMF). 2020. “Toward an Integrated Policy Framework.” IMF Policy Paper 2020/046, Washington, DC. - International Monetary Fund (IMF). 2022. Review of the Institutional View on the Liberalization and Management of Capital Flows. Washington, DC, March. - International Monetary Fund (IMF). 2023a. External Sector Report: External Rebalancing in Turbulent Times. Washington, DC, July. - International Monetary Fund (IMF). 2023b. Global Financial Stability Report: Geopolitics and Financial Fragmentation: Implications for Macro-Financial Stability. Washington, DC, April. - International Monetary Fund (IMF). 2023c. “Guidance Note on the Liberalization and Management of Capital Flows.” IMF Policy Paper 2023/055, Washington, DC. - International Monetary Fund (IMF). 2023d. “Integrated Policy Framework—Principles for the Use of Foreign Exchange Intervention.” IMF Policy Paper 2023/061, Washington, DC. - International Monetary Fund (IMF). 2023e. “Japan: 2023 Article IV Consultation—Press Release; Staff Report; and Statement by the Executive Director for Japan.” IMF Country Report 23/127, Washington, DC. - International Monetary Fund (IMF). 2023f. World Economic Outlook: Fragmentation and Commodity Markets: Vulnerabilities and Risks. Washington, DC, October. - International Monetary Fund (IMF). 2023g. World Economic Outlook: Geoeconomic Fragmentation and FDI. Washington, DC, April. - International Monetary Fund (IMF). 2023h. World Economic Outlook: The Natural Rate of Interest: Drivers and Implications for Policy. Washington, DC, April. - International Monetary Fund (IMF). 2024a. Fiscal Monitor: Putting a Lid on Public Debt. Washington, DC, Oc tober. - International Monetary Fund (IMF). 2024b. Global Financial Stability Report: Cyber Risk: A Growing Concern for Macrofinancial Stability. Washington, DC, April. - International Monetary Fund (IMF). 2024c. World Economic Outlook: Global Prospects and Policies. Washington, DC, April. - International Monetary Fund (IMF). 2024d. World Economic Outlook: Global Prospects and Policies. Washington, DC, October. - International Monetary Fund (IMF). 2025a. “External Positions and Policies.” In External Sector Report: Global Imbalances in a Shifting World. Washington, DC, July. - International Monetary Fund (IMF). 2025b. “The 4th Financing for Development Conference—Contribution of the IMF to the International Financing for Development Agenda.” IMF Policy Paper 2025/022, Washington, DC. - International Monetary Fund (IMF). 2025c. The Global Financial Safety Net—A Stocktaking. Washington, DC, O c to b e r 9. - International Monetary Fund (IMF). 2025d. Global Financial Stability Report: Risk and Resilience in the Global Foreign Exchange Market. Washington, DC, October. - International Monetary Fund (IMF). 2025g. “The Rise of the Silver Economy: Global Implications of Population Aging.” In World Economic Outlook: A Critical Juncture Amid Policy Shifts. Washington, DC. - International Monetary Fund (IMF). 2025h. World Economic Outlook, October 2025: Global Economy in Flux, Prospects Remain Dim. Washington, DC. - International Monetary Fund (IMF). 2026. “Applying the Integrated Policy Framework to Low Income Countries: Shocks, Frictions, and Policy Considerations.” IMF Policy Paper 2026/014, Washington, DC, April. - International Monetary Fund (IMF). Forthcoming-a. Comprehensive Surveillance Review. Washington, DC. - International Monetary Fund (IMF). Forthcoming-b. “Exchange Rate Premia: Drivers and Policy Implications.” IMF Staff Discussion Note, Washington, DC. - International Monetary Fund (IMF). Forthcoming-c. “Integrated Policy Framework: Use of Macroprudential Policies.” IMF Policy Paper, Washington, DC. - Jacks, D. S., C. M. Meissner, and D. Novy. 2008. “Trade Costs, 1870–2000.” American Economic Review 98 (2): 529–34. - Jeanne, O., and R. Rancière. 2011. “The Optimal Level of International Reserves for Emerging Market Countries: A New Formula and Some Applications.” Economic Journal 121 (555): 905–30. - Jiang, W., T. J. Sargent, N. Wang, and J. Yang. 2024. “A p Theory of Taxes and Debt Management.” Journal of Finance, forthcoming. - Kahn, M. E., K. Mohaddes, R. N. C. Ng, M. H. Pesaran, M. Raissi, and J. C. Yang. 2021. “Long-Term Macroeconomic Effects of Climate Change: A Cross-Country Analysis.” Energy Economics 104:105624. - Kalkuhl, M., and L. Wenz. 2020. “The Impact of Climate Conditions on Economic Production: Evidence from a Global Panel of Regions.” Journal of Environmental Economics and Management 103:102360. - Kolasa, M., S. Lassen, and J. Lindé. 2025. “Unconventional Monetary Policies in Small Open Economies.” IMF Working Paper 25/66, International Monetary Fund, Washington, DC. - Kompas, T., V. H. Pham, and T. N. Che. 2018. “The Effects of Climate Change on GDP by Country and the Global Economic Gains from Complying with the Paris Climate Accord.” Earth’s Future 6 (8): 1153–73. - Krishnamurthy, A., and A. Vissing-Jorgensen. 2012. “The Aggregate Demand for Treasury Debt.” Journal of Political Economy 120 (2): 233–67. - Kuzmenko, O., T. Vasylyeva, and V. Bozhenko. 2020. “Modeling of Capital Investments Public Financing Volume Using Logistic Curves.” Proceedings of the International Scientific Conference “Sustainable and Innovative Development in the Digital Age,” Atlantis Press, Paris. - Laubach, T., and J. C. Williams. 2003. “Measuring the Natural Rate of Interest.” Review of Economics and Statistics 85 (4): 1063–70. - Laxton, D., H. Igityan, and S. Mkhatrishvili. 2025. “Monetary Policy Credibility, Avoiding Dark Corners, and Risk Management: A Response to Ben Bernanke’s Review of Monetary Policy-Making at the Bank of England.” Oxford Review of Economic Policy 41 (2): 452–83. - Leeper, E. M. 2013. “Fiscal Limits and Monetary Policy.” NBER Working Paper 18877, National Bureau of Economic Research, Cambridge, MA. - Leeper, E. M. 2023. “Fiscal Dominance: How Worried Should We Be?” Mercatus Policy Brief, April 3. - Levin, A. T., A. Onatski, J. C. Williams, and N. Williams. 2005. “Monetary Policy Under Uncertainty in Micro-Founded Macroeconometric Models.” NBER Macroeconomics Annual 20:229–87. - Macklem, T. 2025. “Flexible Inflation Targeting in a Shock-Prone World.” Speech at the Bank of Mexico’s 100th Anniversary Seminar. - McKibbin, W., and W. L. Liu. 2025. “Long-Term Projections of the World Economy.” CEPR Discussion Paper DP20293, Centre for Economic Policy Research, London. - McKinsey. 2023. “The Economic Potential of Generative AI.” McKinsey Report, January. - Melina, G., and S. Villa. 2025. “From Servers to Rates AI, ICT Capital, and the Natural Rate.” IMF Working Paper 2025/224, International Monetary Fund, Washington, DC. - Mian, A., L. Straub, and A. Sufi. 2025. “A Goldilocks Theory of Fiscal Deficits.” American Economic Review 115 (12): 4253–91. - Misch, F., B. Park, C. Pizzinelli, and G. Sher. 2025. “AI and Productivity in Europe.” IMF Working Paper 2025/067, International Monetary Fund, Washington, DC. - Morris, J., S. K. Rose, J. Reilly, A. Gurgel, S. Paltsev, and C. A. Schlosser. 2025. “Reconciling Widely Varying Estimates of the Global Economic Impacts from Climate Change.” Nature Climate Change 15 (2): 124–27. - Nachega, J.-C. 2005. “Fiscal Dominance and Inflation in the Democratic Republic of the Congo.” IMF Working Paper 2005/221, International Monetary Fund, Washington, DC. - Nakano, S., Y. Sugioka, and H. Yamamoto. 2024. “Recent Developments in Measuring the Natural Rate of Interest.” Bank of Japan Working Paper Series 24-E-12. - Network for Greening the Financial System (NGFS). 2020. “Climate Change and Monetary Policy: Initial Takeaways.” https://www.ngfs.net/en/publications-and-statistics/publications/climate-change-and-monetary-policy-initial-takeaways - Network for Greening the Financial System (NGFS). n.d. “Climate Scenarios Enable Us To.” https://www.ngfs.net/ngfs-scenarios-portal/explore/ - Nguyen, A. D. M., W. R. Lam, and G. Sher. 2025. “Fiscal Rules and Growth-Enhancing Investment: Technical *Executive Summary. Ottawa: Bank of Canada.* ### Annex 3.” In Fiscal Guardrails Against Rising Debt and Looming Spending Pressures, edited by Acalin ### Annex 3. ### Climate, coastal defense, and climate-damage studies - Nicholls, R. J., J. Hinkel, D. Lincke, and T. van der Pol. 2019. “Global Investment Costs for Coastal Defense Through the 21st Century.” Open Knowledge Repository. https://openknowledge.worldbank.org / entities/publication/0dbcb045-3207-5a61-a4ee-e0542c5c0f60 - Tol, R. S. J. 2024. “A Meta-Analysis of the Total Economic Impact of Climate Change.” Energy Policy 185:113922. - Waidelich, P., F. Batibeniz, J. Rising, J. S. Kikstra, and S. I. Seneviratne. 2024. “Climate Damage Projections Beyond Annual Temperature.” Nature Climate Change 14:592–99. - Weitzman, M. L. 2012, “GHG targets as insurance against catastrophic climate damages”, Journal of Public Economic Theory 14(2): 221–44. - Svanberg, M., L. Wensu, M. Fleming, B. Goehring, and N. Thompson. 2024. “Beyond AI Exposure: Which Tasks Are Cost-Effective to Automate with Computer Vision?” January 19. http://dx.doi.org/10.2139/ssrn .4700751 ### AI, labor markets, and productivity studies - Peng, S., E. Kalliamvakou, P. Cihon, and M. Demirer. 2023. “The Impact of AI on Developer Productivity: Evidence from Github Copilot.” Technical Report, arXiv. https://doi.org/10.48550/arXiv.2302.06590 - Pizzinelli, C., A. J. Panton, M. Mendes Tavares, M. Cazzaniga, and L. Li. 2023. “Labor Market Exposure to AI: Cross-country Differences and Distributional Implications.” IMF Working Papers 2023, International Monetary Fund, Washington, DC. - Poplawski-Ribeiro, M. 2020. “Labour Force Ageing and Productivity Growth.” Applied Economics Letters 27 (6): 498–502. https://doi.org/10.1080/13504851.2019.1637509 ### Monetary policy, neutral rate, and financial environment - Orphanides, A., and J. C. Williams. 2002. “Robust Monetary Policy Rules with Unknown Natural Rates.” Brookings Papers on Economic Activity 2002 (2): 63–145. - Platzer, J., and M. Peruffo. 2022. “Natural Rate of Interest in the United States: A Quantitative Evaluation.” IMF Working Paper 2022/030, International Monetary Fund, Washington, DC. - Rachel, Ł., and L. H. Summers. 2019. “On Falling Neutral Real Rates, Fiscal Policy, and the Risk of Secular Stagnation.” NBER Working Paper 2619, National Bureau of Economic Research, Cambridge, MA. - Schnabel, I. 2025. “No Longer Convenient? Safe Asset Abundance and r*.” Keynote Speech, Bank of England 2025 BEAR Conference, London. - Sargent, T. J. 1982. “The Ends of Four Big Inflations.” In Inflation: Causes and Effects, edited by R. E. Hall. Chicago: University of Chicago Press. ### Fiscal policy, public finance, and debt-related analyses - Traum, N., and S. C. Yang. 2015. “When Does Government Debt Crowd Out Investment?” Journal of Applied Econometrics 30 (1): 24–45. - Wu, J. C., and Y. Xie. 2025. “Unconventional Monetary and Fiscal Policy.” Review of Economic Dynamics 5 6:101259. - Reis, R. 2026. “Financial Repression in the XXIst Century.” CEPR Discussion Paper 21072, CEPR Press, Paris. - IMF DEPARTMENTAL PAPERS • Resilience amid Structural Transformations91 - IMF DEPARTMENTAL PAPERS • Resilience amid Structural Transformations92 ### Exchange rate, operational policy, and defense spending references - Ötker-Robe, M. I., and D. Vavra. 2007. “Moving to Greater Exchange Rate Flexibility: Operational Aspects Based on Lessons from Detailed Country Experiences.” IMF Occasional Paper 256, International Monetary Fund, Washington, DC. - Reserve Bank of Australia (RBA). 2022. “Review of the Yield Target.” Sydney. - Schroeder, Wayne. “The pathway to NATO’s ‘2/20’ goal is through real growth of defense spending.” New Atlanticist, Atlantic Council, September 15, 2023. https://www.atlanticcouncil.org/blogs/new -atlanticist/the-pathway-to-natos-2-20-goal-is-through-real-growth-of-defense-spending/ ### Other methodological and thematic contributions - Nordhaus, W. D., and J. Boyer. 2000. “Warming the world: Economic models of global warming.” The MIT Press. - Nordhaus, W. 2014. “Estimates of the social cost of carbon: Concepts and results from DICE-2013R model and alternative approaches.” Journal of the Association of Environmental and Resource Economists 1(1/2): 273 – 312. - Okuda, T., and T. Tsuruga. 2024. “Geoeconomic Fragmentation and International Diversification Benefits.” IMF Working Paper 2024/048, International Monetary Fund, Washington, DC. - Organization for Economic Cooperation and Development (OECD). 2011. “ISIC Rev. 3 Technology Intensity Definition.” OECD Directorate for Science, Technology and Industry, Economic Analysis and Statistics Division. https://www.oecd.org/sti/ind/48350231.pdf - Shapiro, A. H. 2022. “Decomposing Supply and Demand Driven Inflation.” Working Paper 2022-10, Federal Reserve Bank of San Francisco. - Wong, A. 2019. “Refinancing and the Transmission of Monetary Policy to Consumption.” Unpublished Manuscript. https://jfhoude.wiscweb.wisc.edu/wp-content/uploads/sites/769/2019/09/Arlene-Wong -refinancing.pdf ### Publication and identifier details - 042925798229 - 9 - ISBN 9798229042925 - Resilience amid Structural Transformations: Are Macro Policy Frameworks Fit for Purpose? *Annex 3. In Fiscal Guardrails Against Rising Debt and Looming Spending Pressures, edited by Acalin and others. Washington, DC: IMF.* --- _Source: https://www.imf.org/-/media/files/publications/dp/2026/english/rastea.pdf_