The El Niño-Southern Oscillation (ENSO) is a major climate phenomenon characterized by alternating warm (El Niño) and cool (La Niña) phases in the equatorial Pacific Ocean. These phases significantly alter global atmospheric circulation patterns, leading to widespread impacts on weather, hydrological cycles, ecosystems, and agriculture. While the physical dynamics and climate teleconnections of ENSO are relatively well-understood, the precise economic consequences and how these consequences might change in a warming climate remain crucial research areas. Previous studies have documented significant economic downturns in affected regions due to extreme weather events associated with major El Niño and La Niña occurrences. However, these studies primarily focus on direct, tangible losses, often neglecting the longer-term, indirect impacts on the global macroeconomy. This study aims to address this gap by developing a more sophisticated model capable of capturing the nonlinear and persistent effects of ENSO on global economic production. Understanding the magnitude and duration of ENSO's economic impact is critical for effective economic planning, risk management, and the development of climate change adaptation and mitigation strategies. The increasing likelihood of more intense and frequent ENSO events under climate change underscores the urgency of this research.

Existing literature demonstrates a well-established link between ENSO and various aspects of economic production, particularly in sectors like agriculture and fisheries. Studies have shown that El Niño events often lead to reduced crop yields, disrupted fishing patterns, and increased damage from extreme weather, resulting in observable economic losses. However, most prior econometric models treat ENSO's impact as a simple linear predictor, assuming a symmetrical positive impact from La Niña that offsets the negative impact from El Niño. This assumption has been challenged by studies that reveal the asymmetrical nature of ENSO and the potential for La Niña events to also generate negative economic consequences, particularly in the context of extreme rainfall and flooding. Furthermore, existing work often overlooks the delayed, cascading effects of initial ENSO shocks on global economic activity. This study builds upon prior research by incorporating these crucial aspects—nonlinearity, asymmetry, and lagged effects—into a more comprehensive econometric model.

The authors developed a fixed-effect panel regression model using a nonlinear ENSO index (Niño3.4 SST anomaly) as a predictor. The model incorporates lagged effects to account for the persistence of ENSO's impact on economic production over multiple years following an initial shock. The model controls for country-specific fixed effects, long-term linear and quadratic time trends in growth rates, and the nonlinear effects of temperature and precipitation (with the ENSO signal removed to isolate ENSO's independent effect). A quadratic function was employed to represent the nonlinear relationship between ENSO and economic growth, allowing for both linear and nonlinear lagged effects. The model was trained using climate and economic data from 1960 to 2019. A bootstrap method was utilized to quantify the uncertainty of point estimates. The model was further analyzed to explore the heterogeneous effects of ENSO across different groups of countries, including those with strong versus weak ENSO teleconnections, agriculture-dependent versus agriculture-independent countries, and high-income versus lower-income countries. To project future economic losses under climate change, the authors used output from CMIP6 climate models under four Shared Socioeconomic Pathways (SSPs) representing different greenhouse gas emission scenarios. These models project changes in ENSO variability for the 21st century, which are then incorporated into the econometric model to estimate the additional economic losses resulting from increased ENSO amplitude. A counterfactual scenario with unchanged ENSO variability was also constructed to isolate the impact of increased variability. Various discount rates were applied to account for the time value of money when calculating cumulative economic losses.

The model revealed a nonlinear relationship between ENSO and economic impact. Both extreme El Niño and La Niña events cause negative economic growth, but El Niño's impact is substantially larger. Weak to moderate La Niña events show small positive effects which are far outweighed by the damage from El Niño. The impact of El Niño persists for three years after the initial event, which is a significantly longer period than indicated in previous studies. The cumulative global economic losses attributed to the 1997-98 and 2015-16 El Niño events are estimated at US$2.1 trillion and US$3.9 trillion respectively. This is a much higher estimate than that from previous studies that relied solely on direct, tangible losses. In contrast, the cumulative gain from the 1998-99 La Niña event is estimated to be a mere US$0.06 trillion. The analysis of country-level heterogeneity shows that the response to El Niño is larger in teleconnected countries, agriculture-dependent countries, and lower-income countries. Future projections based on CMIP6 climate models under different emission scenarios demonstrate that increased ENSO variability under climate change will lead to significantly higher economic losses. For example, under a high-emission scenario (SSP5-8.5), there is an over 80% chance of an additional median loss of US$33 trillion by the end of the 21st century (at a 3% discount rate). Mitigation efforts, as reflected in lower emission scenarios, would substantially reduce these losses. A strong correlation was found between changes in ENSO variability and changes in global GDP growth rate reduction.

The findings highlight the substantial and persistent economic costs associated with El Niño events, challenging previous estimates that underestimated the true economic burden. The long-lasting negative effects of El Niño, coupled with the increasing likelihood of more intense ENSO events in a warming climate, pose a significant threat to global economic stability. The asymmetry between El Niño and La Niña impacts underscores the importance of considering the full range of ENSO-related risks, rather than assuming an offsetting effect. The study’s findings are particularly relevant for policymakers involved in climate change adaptation and mitigation strategies. The significant economic losses projected under high-emission scenarios highlight the economic imperative of reducing greenhouse gas emissions to limit the future costs of ENSO-related disasters. The significant uncertainty from climate model differences emphasize the need for ongoing research to refine the projections of ENSO variability under various emission scenarios.

This study provides compelling evidence that ENSO events inflict far greater economic damage on the global economy than previously thought, especially as ENSO variability intensifies under climate change. The nonlinearity of ENSO's impact and its persistence over time necessitate a more sophisticated approach to economic modeling. The substantial projected increases in economic loss underscore the economic benefits of aggressive emission reduction strategies to mitigate the devastating impacts of future ENSO events. Future research should focus on refining the economic models to better capture the intricate details of various transmission pathways, exploring the regional heterogeneity of economic impacts more comprehensively, and improving the accuracy of projections on future ENSO variability.

The study acknowledges some limitations. The econometric model relies on certain assumptions regarding the relationship between ENSO and economic variables, and the accuracy of the projections depends on the reliability of climate models and socioeconomic pathways. The uncertainty associated with climate model projections of future ENSO variability is a significant factor affecting the precision of the economic loss estimates. While the study attempted to control for several factors, there might be other unobserved variables influencing the relationship between ENSO and economic growth.