Heatwaves may not just be an environmental problem, but also a behavioural one
In the summer of 2003, a heatwave hit Paris and killed nearly 15,000 people. That same year, temperatures in Dubai regularly climbed past 45 degrees Celsius, and the city carried on much as usual. The crucial difference between the two was that Dubai also had decades of infrastructure and habit built around expecting it.
The same gap shows up in behaviour, not just health statistics. Heatwaves in European cities have been linked to sharper spikes in aggression and disorder, sometimes tipping into open unrest, in places that were never designed with extreme temperature in mind. Cities built for heat tend to fare better. Air conditioning helps, but so do other factors such as work schedules that shift around the hottest hours, and a cultural memory of what to do when the temperature turns dangerous.
Heat clearly affects the mind. What's less obvious, and more interesting, is what happens when temperatures rise drastically in cities that never expected to need a solution for a problem like this.
What heat actually does to the brain
The link between heat and behaviour has a name in psychology: the heat-aggression hypothesis. The basic claim is that rising temperature increases physiological arousal, and that arousal narrows the gap between irritation and reaction.
A few things happen at once when the body overheats. Sleep gets worse, often for several nights running. Cognitive resources that would normally go toward patience and self-control get redirected toward simple thermoregulation, i.e., the body managing its own temperature. Heat also raises baseline stress hormones, which puts people into a more reactive state before anything has even gone wrong.
However, it is important to note that this, by no means, suggests that heat manufactures aggression from nothing. Craig Anderson's General Affective Aggression Model, one of the more influential frameworks in this area, argues that heat works by lowering the threshold for aggression that already exists in a situation, rather than creating hostility out of thin air. A minor disagreement that would normally pass without incident becomes something a person is more likely to escalate, simply because their capacity to absorb friction has been quietly worn down by the temperature around them.
Therefore, heat isn't the cause of unrest. It's closer to an amplifier, and what it amplifies depends heavily on what was already there.
How would behavioural science explain this?
The COM-B model offers a clean way to explain what's actually shifting during a heatwave. Capacity drops, since heat drains the mental resources needed for patience and impulse control. Motivation shifts too, though less obviously. Rising temperature changes how threatening a situation feels, so people become quicker to read ambiguous behaviour as hostile. Opportunity barely changes at all. What does change, however, is a person's ability and inclination to hold back.
Dual-process theory adds another layer to this. Under normal conditions, most people run on a mix of fast, automatic thinking and slower, deliberate thinking, switching between the two depending on the stakes. Heat pushes that balance toward the fast system. Decisions that would usually get a moment of consideration instead get made on instinct, and instinct, under heat, tends to run hotter than it should.
Put together, these two frameworks suggest something quite specific. Heatwaves seem to shift the mental conditions under which decisions get made, and people often misread the resulting arousal as coming from whatever is directly in front of them, and not the temperature that actually produced it.
The gap in infrastructure, and why colder climates suffer more
If heat mainly acts by lowering the threshold for aggression, then the environment someone is standing in when the heat hits starts to matter a great deal. This is the key difference between Paris and Dubai that must be explored further.
Hot-climate cities have spent decades solving for heat without necessarily framing it as a behavioural problem. Air conditioning is the obvious piece. Work and school schedules often shift around the hottest part of the day. Public spaces are built with shade and airflow in mind. Even informal habits, like staying indoors between midday and late afternoon, carry a kind of accumulated wisdom about when not to be outside. None of this was designed with COM-B in mind, but it functions almost exactly like it was. It restores Capacity by protecting sleep and reducing physical strain, and it limits Opportunity by keeping people out of high-friction situations during the hours when tempers are most likely to fray.
Cold-climate cities rarely have any of this. Homes and offices are built to hold heat in instead of releasing it. Public transport systems that are designed for milder weather become unbearable rather than merely uncomfortable. There's no cultural script for what people should do when a heatwave hits, because for most of these places' history, there has never been a need to develop one. So when the temperature spikes, there's no infrastructure absorbing the impact and no accumulated behavioural adaptation stepping in to soften it. Therefore, the effects described earlier, the depleted self-control and the misattributed arousal, play out with nothing standing in the way.
This can, therefore, be framed as a choice architecture story, even though that wasn't designed intentionally. Hot climates have unknowingly built environments that nudge people away from heat-driven conflict. Cold climates haven't, and as heatwaves become more frequent in places like this, that gap is likely to matter more.
The evidence: riots, crime data, and unrest
One of the earliest studies came from Baron and Ransberger, who examined riots in the United States through the 1960s and 1970s and found what became known as the "long hot summer" effect. Riot likelihood rose with temperature, but only up to a point. Past a certain threshold, the relationship curved back down, since extreme heat appears to suppress the physical energy needed for large-scale unrest even as it raises irritability. This curvilinear pattern makes matters interesting because it complicates the simple story of "hotter equals more violent". It points instead to a specific range within which discomfort is high but physical capacity for confrontation has not collapsed yet.
More recent economic research backs this up too. Matthew Ranson's 2014 study in the Journal of Environmental Economics and Management analysed decades of US crime data and found that higher daily temperatures were associated with meaningful increases in violent crime, even after controlling for other seasonal factors. Moreover, the 2011 England riots broke out during an unusually warm August, and while poverty, policing, and a specific triggering event were the clear proximate causes, several commentators at the time noted the heat as a contributing condition rather than the cause itself.
Taken together, heat appears to raise the odds that existing tension turns into something worse.
Looking into the future
Everything so far has treated the Paris and Dubai comparison as a snapshot with two different cities, two different outcomes, and one clear infrastructure gap. But climate change turns this into a bigger problem.
Cold-climate cities are getting hotter, and they're getting hotter faster than the systems around them are adapting. Heatwaves that used to be rare anomalies are becoming a predictable part of summer in cities across northern Europe, and the infrastructure gap described earlier isn't closing at anywhere near the pace the temperatures are rising. London, Berlin, and Toronto are still, for the most part, building and operating as though extreme heat is an exception rather than a recurring feature.
This matters beyond individual discomfort. If the mechanism outlined earlier holds, depleted self-control, misattributed arousal, an environment that offers no relief, then more frequent heatwaves in unprepared cities could mean more of the exact conditions under which minor friction turns into something worse, recurring in places that have no cultural or infrastructural memory of how to absorb it.
The policy implications follow fairly directly from this. Cooling centres and shaded public space are the obvious response, but the behavioural layer matters just as much. Public messaging that names irritability as a heat effect could reduce the chance that people misattribute their arousal to whoever happens to be nearby. Shifted work and school hours during heat events, which is already standard in hot climates, would directly restore some of the Capacity that COM-B suggests gets depleted first. None of this requires new science, only the willingness to treat heat as a behavioural risk and not solely a health one.
As heatwaves keep showing up in places that were never built for them, it becomes increasingly important to notice the gap in time to close it.