Global supply chains amplify economic costs of future extreme heat risk

Research has been showing a trend in rising temperature and increasing occurrence of extreme heatwaves since the 1950s. This continuous pattern raises concerns about the potential impacts of climate change and its associated socioeconomic costs. Notable effects of heat stress are on human health and labour productivity: global heat stress increases morbidity and mortality from heat stroke, and biometeorological studies show serious decreases in labour productivity through lost worktime and reduced work efficiency. In increasingly integrated global supply chains, the impacts of heat stress extend well beyond directly exposed populations and sectors in low latitudes, leading to spillover effects that can affect food security, energy supply and mineral products. Despite extensive study of direct mortality and productivity losses, indirect losses due to supply-chain disruptions have not been fully analysed, as prior work often reports only total effects or overlooks the amplifying effect of global trade. There is a need for methodologies that comprehensively quantify both direct and indirect impacts of heat stress on human systems to inform effective mitigation and adaptation policies. This study constructs an analytical framework to assess the socioeconomic impacts of heat stress to 2060 across 141 regions and 65 sectors worldwide, focusing on health loss (excess mortality), labour productivity loss, and indirect loss (production disruptions propagated via supply chains).

Previous literature has documented rising temperatures, increased frequency and intensity of heatwaves, and their direct impacts on health (morbidity and mortality) and labour productivity. Numerous studies quantify heat-related labour capacity reductions and associated economic burdens, and region-specific analyses have assessed heatwave impacts in Europe, North America, Asia and Africa. However, indirect losses stemming from supply-chain disruptions remain underexplored: prior work often aggregates direct and indirect effects without disentangling them or omits the amplifying role of global trade systems. This study addresses that gap by explicitly modelling how global supply chains propagate and magnify direct heat-related losses into broader economic impacts across regions and sectors.

The study develops a disaster footprint analytical framework integrating climate, epidemiological, and economic trade modules to quantify midcentury heat-stress impacts on socioeconomic systems. Climate inputs: Daily temperature and humidity projections are derived from the Coupled Model Intercomparison Project Phase 6 (CMIP6), averaging 14 global climate models (GCMs), to assess future daily conditions. Epidemiological and labour modules: Grid-scale daily excess mortality (health loss) and labour loss rates (labour productivity loss) are computed using empirical exposure–response functions and statistics from prior studies. Economic trade module: A hybrid input–output (IO) and computable general equilibrium (CGE) global trade model is embedded within the disaster footprint framework to capture how direct losses (mortality and labour productivity constraints) induce indirect economic disruptions through supply chains across 141 regions and 65 sectors. Loss categories: Health loss (excess mortality due to extreme heatwaves), labour productivity loss (decreased daily labour productivity due to higher temperature and humidity), and indirect loss (production stagnation due to supply or demand constraints transmitted via global value chains). Scenarios: Three combined SSP–RCP pathways are assessed—SSP 119 (sustainable pathway with pervasive mitigation), SSP 245 (middle-of-the-road development with continuation of historical mitigation), and SSP 585 (high-emissions, high-growth trajectory). Temporal scope: Estimates are provided as 10-year averages centred on 2040, 2050, and 2060 (for example, 2055–2065 for 2060). Trade relationships are treated as static for estimation, acknowledging that this does not fully capture dynamic industry–country adjustments over the long term. Uncertainty and sensitivity: The study conducts sensitivity analyses on parameters such as stock ratios, excess production capacity, and trade substitutability (with bounds from perfect substitution to no substitution via static IO). Alternative base IO datasets (GTAP 2011 vs GTAP 2014) are compared to assess the robustness of indirect loss estimates, alongside Monte Carlo simulations (10,000 iterations) and historical validation using multiple data sources. Outputs include global and regional value-added (GDP) losses, decomposition by loss type, sectoral impacts, and trade flow changes along representative supply chains.

• Global GDP losses increase nonlinearly with time and heat stress severity, driven by rapidly growing indirect (supply-chain) losses. Under SSP 585, total losses rise from 1.5% of GDP in 2040 to 2.5% in 2050 and 3.9% (2.9–4.5%) in 2060.
• Annual incremental global GDP loss increases exponentially from 0.03 ± 0.01 percentage points (SSP 245)–0.05 ± 0.03 (SSP 585) during 2030–2040 to 0.05 ± 0.01–0.15 ± 0.04 during 2050–2060.
• By 2060, expected global economic losses total 0.6–4.6%, with contributions: health loss 37–45%, labour productivity loss 18–37%, indirect loss 12–43% (varies by SSP).
• Scenario-specific totals in 2060: 0.8% (SSP 119), 2.0% (SSP 245), 3.9% (SSP 585). Indirect loss shares: 0.1% of global GDP (13% of total) under SSP 119; 0.5% (25%) under SSP 245; 1.5% (38%) under SSP 585.
• Under SSP 119: 2040 total global GDP loss 0.9% (0.6–1.1%) composed of 0.5% health, 0.3% labour, 0.1% indirect; by 2060, total slightly decreases to 0.8% (0.4% health, 0.3% labour, 0.1% indirect), ≈US$3.75 trillion (2020 prices). Global average heatwave days increase by 24% vs 2022; heat-induced deaths ≈0.59 million (0.44–0.74 million) annually.
• Under SSP 585 in 2060: total loss 3.9% (2.9–4.5%) comprising 1.6% health, 0.8% labour, 1.5% indirect, ≈US$24.70 (18.36–28.80) trillion. Heatwave days +104% vs 2022; heat-induced deaths ≈1.12 million (0.85–1.39 million) annually.
• Indirect losses accelerate per decade from ~0.1% to 0.3%, 0.7%, and 1.5% of global GDP from 2030 onward (SSP 585), becoming a dominant contributor as heat stress intensifies.
• Spatial patterns: Under SSP 119, health losses highest in South-Central Africa and Eastern Europe; labour losses concentrated in lower latitudes (West Africa, South Asia); indirect losses concentrated in Central America and East Asia; total losses most severe in Central/Southern Africa, Southeast Asia, and Latin America.
• As heat stress intensifies (SSP 585), supply-chain disruptions extend globally: China’s indirect losses rise from 0.4% (SSP 119) to 2.7% of GDP; Brazil 0.2% to 2.5%; Norway <0.1% to 2.1%. EU countries face considerable indirect losses via reduced capacity of trading partners (minerals, food), amplified when many developing partners are simultaneously affected.
• Differential sensitivities: Health losses (SSP 119, 2060) are highest in South-Central Africa (1.8% of GDP), followed by Trinidad and Tobago (1.7%), Sri Lanka (1.5%), Indonesia (1.5%). Labour productivity losses hit low-income, hot-climate economies (e.g., Botswana 1.3%, Nepal 1.2%, Nigeria 1.2%). Indirect losses (SSP 119) are highest in Puerto Rico (0.8%), with 0.4–0.8% in Venezuela, Malaysia, El Salvador, Panama, Dominican Republic.
• Under SSP 585, losses escalate and remain uneven: many African countries (e.g., Malawi, Madagascar, Tanzania) face labour productivity losses ≈2.5–4.0% of GDP; South-Central Africa and Rwanda face health losses of 8.6% and 7.2% of GDP; indirect losses widespread across developed and developing economies (e.g., Brunei 4.7%, Paraguay and Indonesia ≈3.3%).
• Sectoral impacts: Primary sectors (crop farming, construction, mining) are most affected in many African and Asian countries. Under SSP 585 in 2060, indoor manufacturing industries can lose 6.0–7.4% of VA in some countries; in India, ferrous metals lose 5.0% of VA (≥70% indirect), non-metallic manufacturing 3.9%. In Germany, beverages and tobacco rise to 2.0% VA loss and metal products to 2.4% by 2060 (SSP 585), driven by upstream shortages (plantation products, coal, metals). Sweden’s high-end machinery/equipment and chemical sectors see rapidly rising indirect losses under SSP 585 despite cool climate.
• Supply-chain propagation: India’s food sector exhibits an ‘upstream constraint’ (reduced palm oil from Malaysia −5.3% and Indonesia −4.9% by 2060, plus 4–6% declines in other inputs) leading to 3.7–5.1% import contractions in key downstream partners. Dominican Republic tourism shows a ‘downstream constraint’: reduced demand from key markets (e.g., USA tourism −5.5%) propagates to upstream services and manufacturing (business services and manufacturing −4.5–4.7%; reduced demand for US insurance/financial services and Chinese equipment/chemicals by ≈4–6%).

The study demonstrates that global supply chains amplify the economic costs of future heat stress by transmitting and magnifying direct health and labour productivity shocks into broader, nonlinear economic losses. This addresses the research gap on indirect effects by explicitly quantifying how disruptions propagate across interconnected markets and sectors. As heat stress intensifies, indirect losses grow exponentially and account for an increasing share of total GDP loss, shifting from regionally contained impacts under low-to-moderate stress to widespread global disruptions under high-emissions trajectories. The findings reveal heterogeneous vulnerabilities: cooler-climate regions can experience substantial health losses due to abrupt temperature extremes; low-income, hot-climate economies face pronounced labour productivity losses due to the prevalence of outdoor and primary-sector work; and small to medium-sized, trade-dependent economies are particularly exposed to indirect losses through tight linkages with climate-vulnerable suppliers and customers. Sectoral analyses clarify asymmetric impacts: primary sectors suffer directly from heat, while manufacturing and high-end sectors increasingly incur indirect constraints from upstream material shortages or downstream demand contractions. These insights underscore the importance of targeted adaptation strategies, international cooperation among supply-chain stakeholders, and the broader economic benefits of ambitious emissions reductions that curb both direct and indirect losses.

By integrating climate, epidemiological, and global trade models, the study provides a comprehensive quantification of direct (health and labour) and indirect (supply-chain) economic losses from heat stress to midcentury. It shows that indirect losses grow rapidly with heat stress severity and can dominate total losses under high-emission scenarios, with disproportionate burdens on developing countries and trade-dependent economies. The results highlight the value of coordinated global mitigation to limit warming and reduce cascading supply-chain risks, and of targeted adaptation (e.g., technology transfer, sector-specific resilience measures) to protect vulnerable sectors and regions. Future research should refine dynamic modelling of production and trade adjustments, improve representation of product substitutability and stock capacities, and further assess supply-chain resilience as emerging markets in Africa, South America and Southeast Asia deepen their participation in global value chains.

Estimates are based on static production and trade relationships, which may not capture dynamic adjustments among industries and countries over time. Uncertainties exist in parameters such as product substitutability, maximum stock ratios, and excess production capacity. Although sensitivity analyses (including alternative GTAP base years, parameter bounds, and Monte Carlo simulations) indicate global-scale indirect loss estimates are robust (differences <5% by 2060), some regional results vary with trade structure assumptions (e.g., amplified indirect losses in East and Southeast Asia using GTAP 2014). The disaster footprint approach, while widely used, requires further development for multicountry contexts to better address substitutability and dynamic behaviours. Despite these limitations, the overarching conclusions about increasing heat-related risks and amplified economic losses via global supply chains remain robust.