How much will climate change affect the world economy?

Unfortunately, we don’t know. The risks are severe, and grow worse the longer we wait to address them, but climate scientists and economists are very far from agreeing on an exact dollar cost.

February 19, 2026

A world that has warmed 2 or 3°C is not one we would choose to live in. Floods, droughts, and heatwaves would be more extreme, weather risks would threaten food supplies, and rising seas would endanger coastal communities.¹ This would be a more perilous world, and likely a less wealthy one.

But how much less wealthy? To tackle that question, researchers have built models of the world economy in a warming climate—but their results differ enormously.

“There’s just a lot of uncertainty,” says Jennifer Morris, a principal research scientist at the MIT Center for Sustainability Science and Strategy who studies the economics of different climate scenarios. “The estimates are so wide-ranging that they're not really helpful for making decisions.”

This has practical consequences. Policymakers, for example, often assign a “social cost of carbon” to our climate-warming carbon pollution: a dollar amount that each extra ton of this pollution costs the present and future economy. That number can be used to weigh whether specific climate solutions are worth the price, or to set a tax on carbon that accounts for its harms.

Unfortunately, says Morris, “the literature makes it very easy to pick and choose the numbers you want to use.” Published estimates of the social cost of carbon range from $10 per ton or less, all the way up to several thousand dollars.² And governments do pick numbers that are convenient for their priorities: The United States, for example, has pegged the social cost of carbon as high as $190 and as low as $1 at different times under different administrations.

Politics aside, there are valid reasons researchers can reach very different answers about the costs of climate change. One is scientific uncertainty. The climate is highly complex, and we can only predict to within a wide range how much extra warming additional carbon pollution will bring.³ Another is the need to make judgment calls on how much we value the wellbeing of future generations. Is a dollar of benefits to the economy 50 years from now worth a full dollar today?

But even if we could resolve those questions, we would still be left with radical disagreement among the models that try to pin down the costs of global warming.

These models come in two major types, says Morris. The first are “structural”: Scientists combine economic theory and real-world data to build a model of the economy, then prod it to see what happens. Anticipated effects of climate change hit this model economy in specific, traceable ways—“like the impact of temperature on labor productivity, or mortality, or crop yields.”

Morris and her colleagues recently reviewed several prominent structural models. They asked how each one expects climate change to affect the future economy, measured in global GDP per capita. Given 3°C of warming—an amount we could experience by the end of this century—the models’ results ranged from less than 1% loss in GDP to a 5% loss; one model even suggested a chance of up to 3% growth.⁴

“That might sound small, but with global GDP, those are really big numbers,” clarifies Morris. For context, in 2009, the worst year of the Great Recession, global GDP per capita fell by around 2.5%, before rebounding the next year.⁵

Why do the models give such different answers? For one thing, they bake in different assumptions about how climate change will impact each part of the economy. What’s more, they don’t all include the same impacts. For example, Morris says, only recently have some models started including extreme weather events in addition to rising temperatures.

And because climate change alters so much of our world, all these models are necessarily incomplete. “We don't even have estimates for a lot of things, like, how would climate affect biodiversity or ecosystem services, or crime and migration?” she says. “Some of those, like global migration due to mass famine, could be really big. But just how much that would translate into the overall economic impact is still unknown.”

So there is reason to suspect that structural models may undercount the full costs of climate change. But the other major category—“statistical models”—has its own limitations.

These models do not try to parse out climate impacts across a detailed model economy. Instead, they take an empirical approach, searching for relationships between real-world economic and weather data. This data cannot exactly mirror climate change: The world has not warmed by 2 or 3°C in all recorded history. But some years are hotter or rainier than others, especially when looking at smaller regions. Statistical models ask how these variations in temperature and rainfall have affected the economy, then extend the results to possible future climates.

In general, these models expect the costs of climate change to be staggering. Some estimate that 3°C of warming would bring GDP losses over 10, 20, or even 30%,⁶ with at least one model projecting that the world stands to lose over 50% of GDP.⁷ These numbers would make even the Great Depression of the 1920s and 30s seem mild.

“It's hard to even imagine what a 20%, 30%, 60% global GDP loss would be,” says Morris. “I'm skeptical.”

Her biggest question about these models is how well weather data can stand in for climate change. “Weather is very short term, whereas climate is a long-term average,” she says. “And so humans can adapt much better to climate than they can to weather.”

Imagine, for example, that a heatwave hits a local economy hard. Crops fail, workers are less productive, tourists stay away. A model trained on weather data would expect those harms to multiply as the climate gets more extreme. But in the real world, if temperatures keep rising, farmers will plant different crops; businesses will add air conditioning; tourists will visit in spring instead of summer.

“There's a lot of adaptive capacity built into economic systems,” says Morris. “And there's a lot of ability to adapt through global supply chains. If different places are hit worse by climate change, you can shift where and how you produce things.”

Morris hopes that, over time, economists and climate scientists can narrow their estimates of the costs of climate change. In the meantime, acknowledging that there are big open questions does not mean we have to swear off the tools of economics. Policymakers can still look for solutions that deliver the biggest climate benefits for the lowest cost, or set a broadly reasonable carbon tax to nudge people toward cleaner energy and transportation.

“I think it's safe to say that climate change has a negative impact on the economy,” Morris says. “The optimal carbon tax is certainly not zero.”

Most importantly, she warns that uncertainty is a poor excuse to ignore the harms of a warming world. “Even though there is uncertainty, it doesn't justify us doing nothing,” she says. “There's a real possibility of large, potentially catastrophic, climate impacts, and we should be taking action to hedge against those risks before it’s too late.” That means investing now in slowing and eventually stopping climate change, as well as preparing for a warmer and more extreme future.

Submit your own question to Ask MIT Climate

Get the latest from Ask MIT Climate monthly in your inbox

Read more Ask MIT Climate