As politicians and philanthropists shift their attention to other global issues, many journalists and scientists continue to sound the alarm about climate change. There is a tendency, however, to highlight only the most negative interpretations of new research in order to amplify public concern and emphasize urgency. As a result, press releases and news stories far too often overstate or misrepresent the scientific findings they cite. 

A recent report from the Global Carbon Project illustrates this pattern well. The report projects that global carbon emissions from fossil fuels will reach a record high in 2025, an estimated 1.1 percent increase over 2024 levels. Much of the attention has centered on the report’s year-over-year projections of emissions increases. The New York Times reported, “Emissions appear to have stayed nearly flat in China and Europe, but rose significantly in the United States and much of the rest of the world.” The Washington Post went further, claiming that emissions increases are “expected to accelerate in the United States and European Union but slow in China and India” [emphasis added]. 

Setting aside that the Times and Post, citing the same research, conflict on whether Europe’s emissions were flat or rising, the reported findings are understandably concerning to anyone tracking global emissions. And given the steady expansion of renewable energy in the United States, the claim that American emissions are sharply rising seems counterintuitive.

The first important point to note is that year-over-year estimates in this context are not very meaningful. There is no reason to expect emissions declines to occur in a smooth, linear way. They fluctuate in response to short-term factors such as weather and the business cycle. What matters is the long-term trend, and a single year of data cannot tell us whether that trend has changed. As the historical data in Figure 1 show, annual changes in U.S. emissions swing up and down both during periods of long-run growth (1960–2004) and long-run decline (2005–2025).

The more fundamental issue is that the report’s year-to-year estimates are too uncertain to support the strong journalistic claims. The authors are not directly counting emissions; they rely on national energy and industrial statistics with varying degrees of accuracy and feed these inputs into models. Measurement error and modeling uncertainty are unavoidable, and the researchers account for this by showing a range within which the true value likely lies.

The statistical question the report is trying to answer is straightforward: is the estimated year-over-year change large enough to conclude that emission truly rose, rather than remaining the same within the margin of error? If the estimate is not clearly different from zero, then the data cannot distinguish a real change from noise.

Most scientific fields make this determination using a 95 percent confidence level, meaning there’s only about a 5 percent chance that an apparent change stems from random fluctuation. This threshold is long-standing convention, but it is not scientific law and there are settings where a higher or lower standard is more useful. The Global Carbon Project reports its estimates using a much looser 68 percent confidence level.

Even at this lower standard—where the confidence interval is narrower and small effects are more likely to look “real”—the uncertainty range is too wide to distinguish the small year-to-year changes highlighted in the news reports from zero. The data cannot say whether emissions went up, down, or stayed roughly the same.

For the United States, the report’s point estimate is a 1.9 percent increase from 2024 to 2025, but the 68 percent confidence interval ranges from -0.2 percent to +4.1 percent. Because that interval includes zero, the estimate is not statistically distinguishable from no change at all. In practical terms, the data are consistent with the possibility that U.S. fossil fuel emissions did not meaningfully change over the past year.

The same is true for other regions. The report estimates a 0.4 percent change (-0.9 to +2.0 percent) for China, 1.4 percent (-0.3 to +3.1 percent) for India, and 0.4 percent (-2.1 to 2.8 percent) for the European Union. None of these year-to-year changes are statistically significant, even at the relatively low 68 percent confidence level. Globally, the report estimates a 1.1 percent increase (0.2 to 2.2 percent). Although this narrower interval excludes zero, the corresponding 95 percent interval (which the authors do not report but can be approximated) would range from roughly -0.9 percent to +3.1 percent. That means the global estimate is statistically significant at the 68 percent level but not the conventional 95 percent threshold.

Taken together, the year-to-year numbers are at best described as suggestive of an increase in fossil fuel emissions. They do not show a statistically reliable increase in emissions and do not support claims that emissions are “accelerating.”

So, what does the report show conclusively? At a 95 percent confidence level, it finds that fossil fuel emissions declined in 35 countries that experienced economic growth between 2015 and 2024. And the longer-term regional trends, as shown in Figure 2, tell a different story from the news reports. Globally, average annual emissions from 2015 to 2024 were 11 percent higher than in the decade before. However, emissions in the United States and Europe declined by 11 percent and 18 percent, respectively. In contrast, emissions increased by 60 percent in India, 32 percent in China, and 10 percent in the rest of the world during that same period.

All in all, these findings are not very alarming. Interpreted optimistically rather than from a doomsday perspective, the report’s actual conclusions point to stagnant but not reversing emissions trends in advanced economies, signs of slowing emissions growth in developing ones, and evidence that economic growth has been decoupled from fossil fuel emissions in 35 countries. This is a far cry from the journalistic focus on “record highs” or purportedly “accelerating” emissions growth.

The instinct to treat every data point as evidence of a crisis is not useful. It obscures areas of real progress, prevents an honest assessment of climate risk, and makes it harder to balance the costs of mitigation and adaptation with other social priorities. Effective climate and energy policy depends on getting the fundamental facts right.

The views and opinions expressed are those of the author’s and do not necessarily reflect the official policy or position of C3.