Corn ethanol was supposed to help the climate. Instead, its production may have made things worse.

That’s what Monday’s study suggests may have happened: “The previously underestimated emissions from U.S. land conversion attributable to the policy are enough to fully negate or even reverse any [greenhouse gas] advantages of the fuel relative to gasoline,” the authors wrote.

In the study, researchers used economic modeling to calculate the impact RFS had on corn prices, finding that the program raised them by nearly 30 percent. Next, they calculated the relationship between the higher prices and observed land use changes, which include both the expansion of cropland, the replacement of other crops with corn, and declining participation in land conservation programs.

In total, increased emissions from land use changes equalled more than 438 million tons over eight years, equivalent to adding 86 million passenger vehicles to the road for a year. In addition to land use changes, Lark’s team also considered emissions associated with increased fertilizer use, another practice that releases greenhouse gases. Amortized across 30 years (for consistency with how carbon intensity is typically calculated), the team found that corn ethanol’s carbon intensity may actually be 24 percent higher than gasoline.

“No one anticipated such a large response and expansion of crop land,” said Tyler Lark, lead author of the paper and scientist at the University of Wisconsin-Madison Center for Sustainability and the Global Environment. “It’s enough to turn the balance from corn ethanol having net benefits relative to gasoline to being a net emitter relative to gasoline. In other words, it changes it from helping combat climate change to being another cause and contributing factor, likely exacerbating emissions from transportation.”

Climate researchers and environmental advocates have long warned that this might happen. In 2008, a team led by Timothy Searchinger, a scholar at Princeton University’s Center for Policy Research on Energy and the Environment, published a widely read analysis predicting that global land use changes from corn ethanol production could double greenhouse gas emissions.

But in 2010, the Environment Protection Agency (EPA), which administers RFS, determined otherwise. By its own math, after considering emissions from land use change, the agency projected that corn ethanol had a carbon footprint 21 percent lower than gasoline. This number is important—under federal law, renewable fuels are RFS-eligible only if they can achieve a 20 percent greenhouse gas emissions reduction relative to gasoline. 

In other words, EPA’s calculations produced a score for corn ethanol right on the knife’s edge of what was required by law in order for the product to qualify for the program. (If we want to really get into the weeds: What EPA actually found in 2010 was that by 2022, over a decade into the future, assuming advancements in technology and productivity that hadn’t yet been realized, corn ethanol could eventually achieve a carbon intensity 21 percent lower than gasoline. In other words, a moonshot view of things. Apparently, this passed muster, and corn ethanol has been booming in the Midwest since.) In an email, an EPA spokesperson said that the agency was reviewing Monday’s study.

For the corn industry, EPA’s 21 percent calculation was not only a huge win, but also a promise fulfilled by then-President Obama, a vocal advocate for ethanol during the 2008 election. The White House is “sending a very positive, very specific, very direct message that the Obama-Biden administration is highly supportive of the biofuels industry,” then-Secretary of Agriculture Tom Vilsack said at the time, according to an InsideClimateNews report. (Vilsack reprised his role for the Biden Administration.) The EPA under Obama denied that industry influence had affected its assessment.

Why are EPA’s numbers so different from those of the study’s? At the core of the issue is that modeling the impacts of any climate policy is inherently uncertain. What one model projects about the future can be vastly divergent from what ultimately plays out in reality.

“There are a number of assumptions you have to make,” said Richard Plevin, consultant and retired UC Berkeley researcher specializing in biofuel modeling, who reviewed the study. “How fast is technology going to improve so that we get more ethanol out of a quantity of corn? How much is the yield of corn going to improve over time? Can it actually backslide because of climate change and extreme weather events? We don’t know—and that’s the essence of the uncertainty. Any precise number coming out of the model represents just one set of assumptions. None of these models produce ‘truth.’”

The numbers drawn from Monday’s study, too, are based on models, meaning that they, too, are based on assumptions, and shouldn’t be taken as a final say on the carbon footprint of corn ethanol. But what distinguishes Lark’s methodology from EPA’s 2010 analysis is the benefit of hindsight. While EPA’s 2010 projects were based on what it expected to happen in the future, Lark’s analysis is based on land use changes that actually occurred between 2008 and 2016.

In fact, the study notes that it likely undercounts total land use change impacts, pointing out that it doesn’t look at impacts outside of the United States, nor does it include additional emissions resulting from increased fertilizer production. Still, the findings are likely to bolster concerns that the carbon footprint of corn ethanol is underestimated, and that its production may worsen the very issue it was supposed to mitigate. The study also poses a timely question for policymakers going forward: Given the uncertainty around corn ethanol’s climate impact, should it continue to get propped up?

It’s a question that EPA will have to answer in this year, when it’s expected to issue new blending requirements for 2023 and beyond.