Climate Cost Study Pulled After Errors Found, Sparking Fresh Attacks on Climate Science

A widely cited study on the catastrophic economic costs of climate change has been retracted after researchers admitted to significant data errors that overstated the paper’s headline findings.

Published in Nature in April last year, the study — accessed more than 300,000 times and cited by outlets including Forbes and Reuters — estimated the global economic burden of climate damage through mid-century. But on 3 December, the Potsdam Institute for Climate Impact Research (PIK), which led the work, acknowledged that mistakes in the underlying data made the inaccuracies “too substantial” to correct. It is the sixth retraction from Nature this year.

What the study got wrong

The original paper projected a 19% decline in global income by 2050 due to climate change. Updated analysis now lowers that figure to 17%.

It also previously claimed a 99% probability that climate damages by mid-century would exceed the cost of building resilience. That likelihood has dropped to 91% in the revised, not-yet–peer-reviewed version.

The most viral finding — that climate change would cost $38 trillion (€32.5 trillion) annually by 2049 — has been revised down to $32 trillion (€27.4 trillion). PIK says the downgrade stems from the unequal distribution of climate losses: poorer regions suffer proportionally more damage, which reduces the global dollar total.

The correction also means annual climate damages in 2050 are now estimated to be five times higher than the cost of limiting global warming to 2°C, rather than six times higher as previously claimed.

Fuel for climate deniers

The retraction quickly sparked an online wave of conspiracy theories, with some claiming the incident proves climate science is “corrupt” or climate change a “scam.” These unfounded narratives spread despite mainstream coverage from AP News, The New York Times, The Wall Street Journal, and Sky News.

While the study was referenced by the World Bank and other institutions in climate-risk scenarios, there is no evidence it was manipulated to influence markets — a claim circulating widely on social media.

What went wrong in the analysis?

The authors used historical data to estimate how temperature and rainfall affect long-term economic growth. But they later discovered serious flaws in economic data from Uzbekistan between 1995 and 1999, which disproportionately skewed the global results. They also acknowledged that their statistical model underestimated uncertainty, making the estimates appear more precise than they were.

The revised analysis corrects the faulty data, applies stronger controls against anomalies, and accounts for correlations across regions.

‘Core findings still hold’

Despite the retraction, PIK maintains that the central message of the research remains intact: climate damages through mid-century will be “substantial” and will exceed the costs of mitigation.

“These impacts are driven mainly by rising temperatures and disproportionately affect low-income regions with low historical emissions,” the institute said, noting the findings remain consistent with broader climate-economics research.

Climate economist Gernot Wagner, who was not involved in the study, told AP News that the overall conclusion remains unchanged regardless of whether the true cost falls at the high or low end of current estimates.

“Climate change already hits home,” Wagner said, pointing to sharply rising home-insurance premiums in the U.S. over the last decade. “Rapidly accumulating climate risks will only make the numbers go up even more.”