New Economic Methodology Finds Climate Damages Overestimated by Half Without Accounting for Global Convergence

In 2015, a landmark climate-econometric model projected that unmitigated global warming would reduce average global incomes by a staggering 20% by the end of the century. That figure, which also forecast a drastic 118% increase in intercountry income inequality, quickly became a cornerstone of worst-case climate economic scenarios. It suggested a future where the economic divide between the global north and south would become insurmountable.

But a new methodological review published in Energy Economics suggests that those empirical models missed a fundamental macroeconomic mechanism: global growth convergence. When poorer countries experience economic shocks, they tend to grow faster afterward to catch up with richer nations. By ignoring this "convergence" effect, previous models effectively assumed that any climate-induced slowdown in economic growth was permanent, compounding year after year into catastrophic, unrecoverable long-term losses.[1][2]

A research team led by Anthony Harding at the Georgia Institute of Technology, alongside economists from the University of Waterloo and the Kiel Institute for the World Economy, set out to re-estimate the global non-linear relationship between temperature and country-level economic growth. By explicitly accounting for income growth convergence in their empirical investigation, their findings cut the projected global income losses by more than half.

Under a high-emissions scenario, the projected reduction in global average incomes drops from 20% down to 8.5% by 2100. This massive revision fundamentally alters the baseline understanding of how a warming planet will impact the global economy over the next eight decades, providing a more grounded estimate for policymakers calculating the social cost of carbon.[1][2]

The mechanism behind this revision lies in how economies naturally recover from temperature shocks. When temperatures increase beyond optimal levels, economic growth demonstrably slows down, affecting everything from agricultural yields to labor productivity. However, as growth slows, the convergence effect pushes back. Because poorer or temporarily stunted economies have more room to adopt existing technologies and improve efficiencies, they naturally tend to grow faster than mature economies.

This offsetting force increases growth rates as the affected country attempts to return to its baseline economic trajectory. By capturing this dynamic, the new methodology shows that economies recover a significant portion of the economic losses that previous models assumed were permanently erased from the global ledger.[1][3]

Without accounting for convergence, empirical models produce a counterintuitive and mathematically problematic long-term result. They predict exponential benefits for "winning" countries—typically large northern nations like Canada and Russia, where warmer temperatures might temporarily boost agricultural or economic activity. In these older models, the exponential growth of the winners eventually outpaces the losses incurred by "losing" countries closer to the equator.

Because a losing country's economy can only shrink to zero, the unchecked exponential growth of the winners eventually causes the model to suggest that climate change has a net economic benefit in the very long run—a conclusion that contradicts physical climate science and basic planetary limits.[1][2]

By building convergence into the model, the researchers found that the aggregated damages become convex. The model consistently pushes countries back together, preventing the winners from infinitely outpacing the losers. As a result, the revised projections show that climate damages continuously increase over time with ongoing warming, eliminating the paradoxical long-term benefit. The convergence-consistent approach ensures that the economic models reflect a reality where the global economy is interconnected, and where the extreme compounding of both localized benefits and localized losses is naturally moderated by macroeconomic forces.[1][2][3]

This methodological correction also radically alters projections of global inequality, a critical metric for international climate policy and adaptation funding. The previous models suggested that climate change would cause a 118% increase in the Gini coefficient—a standard measure of inequality—by 2100, painting a picture of a world violently fractured by temperature disparities.

The convergence-consistent model reduces that projected increase to just 8%. This indicates that while climate change will undoubtedly cause harm, it has far less impact on inter-country income inequality than previously feared. The damage is spread more evenly across all countries rather than being concentrated entirely in the developing world.[1][2]

Despite the significant downward revision in projected damages, the evidence remains clear that climate change will inflict substantial economic harm. An 8.5% reduction in global average incomes by 2100 still represents tens of trillions of dollars in lost economic output and requires immediate, aggressive mitigation efforts.

The researchers emphasize that while convergence moderates the effects relative to older empirical models, the absolute scale of the damage remains large. The findings highlight a persistent tension in climate economics between backward-looking empirical data and forward-looking projections. In historical data, convergence often doesn't show up as a statistically meaningful predictor of the relationship between temperature and short-term economic growth.[1][3]

Ultimately, this methodology provides a crucial bridge between two previously conflicting schools of thought in climate economics. For years, researchers have debated the gap between the climate-econometric approach, which relies on historical data and often projects massive damages, and Integrated Assessment Models (IAMs), which rely on macroeconomic theory and typically project more modest damages in the range of 6% to 7%.

By incorporating a core macroeconomic principle into a data-driven empirical model, the new projections reconcile this glaring discrepancy. The resulting 8.5% damage estimate aligns much more closely with the theoretical models while remaining grounded in observed historical data, offering a transparent foundation for future climate policy.[1][3][4]