Worldwide Job Losses Due to Natural Hazards (Washington, DC: World Bank , 2026-05-29 ) Isah, Abdulrasheed; Rentschler, Jun; Middelanis, Robin; Avner, Paolo; Hallegatte, Stephane Natural hazards can profoundly disrupt economies, yet their impact on employment remains underexplored. This study quantifies job losses due to floods, earthquakes, wind, storm surges, tsunamis, and heat across 132 countries, using a full-time job equivalent loss estimation approach. The results show that fast-onset natural shocks cause 9.4 million job equivalent losses annually on average, predominantly due to earthquakes and floods, with burdens concentrated in East Asia and the Pacific and Sub-Saharan Africa. Additionally, extreme heat was associated with 79.7 million job equivalent losses annually across 114 countries between 2015 and 2024, with the burdens concentrated in South Asia and Sub-Saharan Africa. Yet, average annual job losses can be significantly lower than losses from specific extreme events, for instance, with 1-in-100-year hazard events resulting in losses that exceed average annual job losses by a factor of over 10. Overall, low-income countries experience the highest job loss rate per capita. Within countries, the poorest population group bears a disproportionate share of job equivalent losses. Results highlight the urgent need for targeted adaptation and resilience measures that safeguard workers, jobs, and productivity to support economic development.
Assessing The Real-World Economic Value of Weather Forecasts under Compounding Extremes: A Decision-Specific Framework (Washington, DC: World Bank , 2026-06-02 ) Olivetti, Leonardo; Messori, Gabriele; Avner, Paolo; Hallegatte, Stephane Assessing the real-world economic value of weather forecasts remains challenging, particularly in the context of high-impact extreme events. Although meteorological skill has improved substantially in recent years—driven by steady advances in physics-based models and impressive breakthroughs in artificial intelligence-based forecasting—operational evaluations still focus primarily on standard skill metrics, with limited consideration of how improvements in meteorological skill translate into economic value. This study proposes a flexible framework to assess the economic value of weather forecasts, with penalty functions that explicitly account for compounding losses as well as declining user trust in cases of repeated false alarms. In addition, the framework allows for varying cost—loss ratios to represent heterogeneous prevention costs and vulnerability structures. The framework is applied to cities exposed to weather-related natural hazards, comparing the relative economic value of leading physics-based and data-driven forecasting systems from the European Centre for Medium-Range Weather Forecasts. The value of forecasts is highly sensitive to assumptions about compounding losses, penalty structures, and prevention costs, which often substantially alter conclusions drawn from meteorological skill alone. For instance, in some cities in Southern Europe, the higher sensitivity of the physics-based Integrated Forecast System high-resolution model (IFS HRES) makes it better suited when protection costs are small relative to potential losses, while the higher specificity of the data-driven Artificial Intelligence Forecasting System (AIFS) makes it better when protection costs are higher. These findings underscore the importance of evaluating economic value under realistic risk scenarios to ensure that improvements in predictive accuracy translate into meaningful societal and economic benefits.
One-Fifth of the World’s Population Is at High Risk of Climate-Related Hazards (Washington, DC: World Bank , 2026-05-15 ) Hill, Ruth; Gorgulu, Nisan; Brunckhorst, Ben; Nguyen, Minh Cong; Hallegatte, Stephane; Naikal, Esther Climate change is increasing the frequency and intensity of extreme weather events. When these events occur, they threaten lives and livelihoods. This paper estimates the global population at high risk of climate-related hazards by combining household-level vulnerability data with local exposure to four types of events: agricultural droughts in rural areas, floods, heatwaves, and cyclones. Under current climate conditions, 4.5 billion people are expected to experience these hazards at intensities exceeding hazard-specific thresholds chosen to capture the likely occurrence of systemic impacts within a lifetime. One-third of this population is considered highly vulnerable, based on seven dimensions that influence ability to cope and recover: income, education, access to finance, social protection, drinking water, electricity, and access to services and markets. Overall, one in five people globally are considered at high risk, meaning they are both likely to experience at least one of these hazards and face severely limited capacity to recover from their impacts. Although the share of the global population at high risk has nearly halved since 2010 due to decreased vulnerability, the number of people exposed has increased, and progress has been uneven across regions. This study introduces a new global population headcount indicator based on household survey data and high-resolution spatial data to monitor climate risks across countries and over time.
A Theory of “Political Will” for Reforms (Washington, DC: World Bank , 2026-06-01 ) Karpuska, Laura; Khemani, Stuti Do politicians pursue reforms when they know that status-quo policies are inefficient? Prior literature has answered this question as dependent upon reelection incentives to pander to uninformed voters. This paper shows that reform failure can arise from political selection, or the intrinsic characteristics of those who pursue leadership positions and careers in politics, even when voters would like governments to pursue efficient policies. First, the paper shows how incrementally changing the environment in pandering models can yield political selection of “status-quo” types rather than “reformers” as the driving force behind reform failures. In stark contrast to prevailing explanations, reelection incentives in this model push politicians toward reforms. Second, however, even when reelection incentives are introduced, “status-quo” types continue to resist reforms because of the reputation value in political markets of sticking to their preferred (inefficient) policy to signal their type. That is, politicians can choose inefficient policies even at the risk of losing elections because of political career concerns after exiting office. The paper thus formalizes the notion of “political will” as intrinsic motivation and political career concerns of different types of politicians. The relevance of the model is illustrated using available case studies and empirical evidence, with forward-looking implications for international development agencies that advocate reforms.
The Cycle of Hate, and What We Can Do About It (Washington, DC: World Bank , 2026-02-09 ) Esses, Victoria; Harell, Allison; Lowe, Matt; Markus, Hazel Rose; Mills, Aaron; Rao, Manasi; Rao, Vijayendra; Reicher, Stephen; Singh, Prerna Intergroup hate—both shaped by and shaping development processes—is spreading worldwide as hate speech becomes normalized, hate groups proliferate, and political discourse increasingly frames opponents as enemies rather than as partners in compromise. Drawing on historical, economic, political, and social-psychological research, this paper synthesizes 10 drivers of intergroup hate into four interlocking components: history, current context, call to arms, and justification of mistreatment. These components form a self-reinforcing cycle that escalates animosity and legitimizes harm, making hate difficult—but not impossible—to disrupt. The paper shows how the 10 drivers interact over time and uses the cycle of hate framework to organize evidence from experiments and program evaluations aimed at reducing intergroup animosity. This evidence indicates that intergroup hate can be interrupted at multiple points through coordinated psychosocial, institutional, and economic interventions. By contrast, policies that neglect any of the four components—particularly elite and media mobilization—consistently underperform. Context sensitive, integrated, institutionally embedded strategies hold the greatest promise, including the potential to support inoculation and early-warning systems that detect and counter intergroup hate before it is politically mobilized.