Should you go to work during a heat wave? Your productivity suffers, and GDP tanks when it’s hot

Productivity peaks at around 22°C (72°F) and falls roughly 2% per degree above 25°C, reaching 8.9% below peak at 30°C.

Europe is melting: a record-breaking heat wave has been crushing the continent since late May, with Paris hitting 38°C, London 33°C, and Berlin matching it. Thousands of deaths are projected before it breaks. And while the images of crowded fountains and shuttered schools dominate the headlines, the economic damage is already accumulating in ways most employers and policymakers have barely begun to reckon with.

In fact, if the heat waves keep worsening, Allianz has estimated that France, Italy, Germany, and Spain alone could absorb cumulative heat-related GDP losses of $638 billion by 2030, driven primarily by falling labor productivity and surging cooling costs.

“Heat affects everything from students taking important exams to workers,” said R. Jisung Park, a labor economist at the University of Pennsylvania’s Wharton School and the author of Slow Burn: The Hidden Costs of a Warming World. “Industrial accidents. You name it.”

The results of heat on the economy are seen everywhere. New York City public school students who take their high school exit exams on a 90-degree day are historically about 10% less likely to pass than they would be on a 65-degree day.

Workers in both indoor and outdoor settings are between 5% and 45% more likely to experience a significant injury or accident on a day in the high 80s Fahrenheit or above. And even in the United States, one of the most heavily air-conditioned countries on the planet, hotter-than-average years produce lower-than-average learning and human capital accumulation among American students, showing up in PSAT scores in what researchers describe as a highly unequal effect.

It’s hot everywhere

For years, the relationship between heat and economic output was more intuition than data, which Park notes dates back to at least Aristotle, and most recently to Singapore’s founding Prime Minister Lee Kuan Yew, who credited air conditioning as the secret behind his tropical city-state’s economic miracle.

“We now know that not only are hotter places much poorer on average—we’ve known that for a while—but that hotter-than-average years lead to lower-than-average GDP growth and output generally in many settings,” Park said.

Canonical research finds that for the world as a whole, one degree Celsius above the average reduces average GDP per capita growth between 0.7% and 1.3%.

To be sure, the effect is uneven as some very cold places may see modest benefits from marginal warming, at least up to a point. But for most of the world, and particularly for countries and companies caught unprepared, the impact on GDP is clearly downward.

Park said the most economically damaging heat is not the record-breaking kind but the gradual ones. “Heat gets a lot of attention when you see these 100-degree, record-breaking heat waves, as we just saw in Europe,” he said. “But the data actually suggests that the bulk of the damage is actually being done in hidden ways during the less extreme events that occur with far greater frequency. They just don’t seem to be on our radar.”

Days in the high 70s and low 80s Fahrenheit are where the accumulated drag lives. “Even days in the 80s have subtle but measurable effects on things like worker productivity, student performance, and workplace accidents,” he said. “These things can add up because there are far more of those days.”

Creating behavioral change

This is less about climate and more about infrastructure. Park referenced how a 90-degree day in Seattle kills far more people than a 90-degree day in Houston, not because the heat is worse, but because Houston has built its entire physical and institutional world around the expectation of it. “The Houstons of the world are just much better adapted to heat,” Park said.

What specifically helps (home AC, workplace AC, heat warning systems, healthcare infrastructure, transportation) remains a subject of active research. “There’s some combination of things that makes places like Houston much more adapted to heat, and makes much of Western Europe not very well adapted to heat, in part because there hasn’t been that much of it,” he said. Germany, France, the Netherlands, and the UK have historically averaged single-digit numbers of days above 90°F per year. Houston has at least 100. “Now it’s creeping up,” Park said.

Between 80% and 90% of Americans have home air conditioning, according to U.S. Energy Information Administration data. In Germany, that figure sits at roughly 19%—nearly double what it was just two years ago, as the country rapidly installs cooling equipment—while in the UK, only around 5% to 7% of homes have AC, per IEA data.

A new European Central Bank working paper found that a single extreme heat day reduces German GDP growth by 0.2 to 0.3 percentage points over the following 12 months, while the effect in more heat-acclimatized Spain and Italy is comparatively muted.

A 2006 Lawrence Berkeley National Laboratory meta-analysis found that productivity peaks around 22°C (72°F) and falls roughly 2% per degree above 25°C.

“You can imagine what happens to countries like Germany, Sweden, UK, on days where temperatures are in the 30s Celsius, and you still have to do the work you need to do, and you don’t have as much air conditioning,” Park said.

A slowdown on workplace productivity

For companies specifically, Park said the heat exposure risk should be a significant conversation. “I’d be curious how many companies could tell you right off the bat what percentage of their workforce is or is not exposed to heat on a daily basis,” he said. “Have they done that assessment? Some companies probably have, but many may have not. Depending on what that percentage is, your bottom line may be affected more or less.”

There’s the direct hit to worker productivity just as there is to the logistics chain. “There’s really good evidence that the efficiency of your logistical operations may be affected by heat,” Park said, referencing flight delays. Research shows that heat increases both delays and cancellations at major airports. That’s mostly human: baggage handlers, ground crew, refueling staff, all workers who are more likely to call in sick, more likely to be injured, and more likely to slow down on a hot tarmac. There’s also a physical threshold above which runways themselves become inoperable. “Some airports are better adapted than others,” Park noted, “which suggests that there’s something that we can do about it.”

Obviously this effect is prominently felt in a massively globalizing economy that relies on transportation and logistics. “Heat may actually be a significant component of losses that you’re not really aware is linked to heat,” Park said. “Maybe driven in part by these kinds of disruptions that kind of add up. Drip, drip, drip.”

There’s also the issue of white-collar workers in well-air-conditioned office buildings while workers in kitchens, warehouses, and on construction sites absorb the full heat load. “There’s a lot of heterogeneity, there’s a lot of inequity,” he said. “White collar workers tend to be affected differently, but as we’ve seen in Europe, depending on what the infrastructure is, white collar workers are not immune either.”

Today, there are European governments discouraging AC use, power grids ill-equipped to absorb new cooling demand, and a growing tension between climate goals and adaptation needs. “This is a case in point of the perils of climate change being so often discussed as an ideological, political issue rather than a pragmatic, practical economic issue,” he said. “What happens is that we mix up all these ideologies and political stances when there’s arguably a more pragmatic middle ground.”

“Let’s be dispassionate, data-driven, and decision-relevant. What the data says is that if you care about livelihoods, if you care about preserving human life, if you care about economic productivity, then be mindful of at least the potential for extreme temperatures to affect all these human outcomes.” His recommendation? “Start with what the data tells us before we overlay ideology.”

“Change can be made in response to a dramatic disaster, or change can be made proactively,” he said. “I would hope that we could do the latter by engaging in a perspective shift, not only categorizing climate change as either a distant environmental problem or an ideological political problem, but also thinking of it as a very present, practical economic problem that already affects our day to day.”

This story was originally featured on Fortune.com