Tempers May Flare and Conflicts Rise as Climate Change Heats Up, Study Finds

A new study shows that changes in the climate also affect human behavior. According to researchers, even slight increases in temperature and rainfall correlate to increased levels of violence. Ray Suarez discusses the findings with study lead author Solomon Hsiang of the University of California, Berkeley.

Read the Full Transcript


    And now: reading a thermostat to predict a violent climate.

    Ray Suarez has the story.


    Hot temperatures and changes in the climate are strongly linked to human violence. That's the conclusion of a new study recently published in the journal "Science." Researchers at Princeton and the University of California at Berkeley pored through data compiled in 60 studies from a variety of disciplines.

    They found that even modest increases in normal temperatures or changes in rainfall increased conflict on all levels, both in ancient times and our own, in wealthy societies and developing ones.

    Solomon Hsiang is the lead author of the study, and he joins me now.

    Professor, I think we have become accustomed in recent years to observing how weather changes natural systems, whether it's waterways or migratory bird patterns or insect activity. But how do you measure, how do you correlate human behavior to temperature?

    SOLOMON HSIANG, University of California, Berkeley: So, in recent years, a variety of research groups have assembled really useful data sets, where we have records of how many conflicts occurred in different locations.

    In some cases, we're just looking at things like the FBI files that record how many assaults or rapes or murders happen in a U.S. county on a given day. And then what we do is we link that kind of data to historical data on the climate, so temperature, rainfall. And we try to understand how a change in the environment can lead to human response on the ground.


    So, to see if I understand this correctly, you're measuring a lot of different kinds of human response. But is it just the temperature that could be forcing these changes? I mean, the temperature may have an effect on how much drinkable water there is around. Temperature may have an effect on how much food there is to eat in a given community.



    So there are in fact many hypotheses, many mechanisms that people think might help connect changes in the environment, changes in the temperature or extreme rainfall to the conflict outcomes that we observe. Sometimes, we think direct exposure to heat sometimes actually does change human psychology.

    We observe, even in a laboratory, if we put people in a room and raise the temperature, they actually change how they behave towards others. But, as you suggest, there's economic mechanisms as well that are incredibly important. So you can have crop failures when you have extremely high temperatures, and that leads to all sorts of changes.

    It changes people's incentives to participate in the formal labor market, in comparison to more violent activities. And it also changes how people migrate, food prices. And all of those things can have an influence on human conflict, which is a very complex phenomena and is affected by both how we…


    Is there a temperature or an amount of increase where you see the effects really intensify or start to take off? Is there a threshold where you really see your observable results more strongly?


    So it turns out that, when we look at temperature, it actually looks as though the relationship is pretty continuous.

    So, pretty much wherever you are in the modern world, we tend to observe that even increasing the temperature by a little bit leads to higher levels of conflict.

    But, if we look at rainfall data, there does tend to be these sort of threshold effects, where very extremely high levels of rainfall or extremely low levels of rainfall tend to be damaging. And that's consistent with the economic ideas you have been pointing out, particularly because extreme rainfall is very bad for agriculture.


    It's interesting that you tried to do this retrospectively.

    Is there reliable enough data both on human activity and on the natural world to look back into the past and say, aha, here's a time where rising temperatures can be correlated with increased human violence?


    Data quality is always an issue whenever you do any sort of retrospective study.

    We looked across 60 studies, and some of them go way back in time, thousands of years. But roughly half the studies come from the modern era. So, these are studies — these are studies of data where the populations that are being observed are populations from 1980 to the present.

    And so, in those situations, we think the data is of very high quality. And we are able to actually observe strong associations like you suggest.


    Are there examples of the opposite, places that literally cool off when they cool off?


    So, yes.

    So, when we look at these sorts of relationships, what we're actually saying — we — is that higher temperatures tend to be worse, cooler temperatures tend to be better in the modern world. But if we go back far enough in time — so, if you were to look in Europe during the Little Ice Age, for example, you would actually see that cold events, during these cold epochs in cold locations, also tended to lead to more conflict between populations.

    And that seems to be a general pattern through many studies that look back in time far enough. In the modern world, we're obviously not in that kind of ice age, and we tend to see that higher temperatures are worse.


    Conflict is a pretty big word. And I suspect a lot of things could live under that broad umbrella.

    What kind of effects were you looking for when you were looking for things to measure? What are the — was it just the examples of assaults, of crimes, of murders? What exactly is conflict?


    So, when — we were trying to look at all different types of human conflict. But to sort of organize it and help us understand better what are the general patterns, we broke conflict down into three type of conflicts.

    So, we look at interpersonal conflict — that's sort of conflict between individuals, things like murder, assault, rape, domestic violence. Then, we also looked at intergroup conflict. So, that is conflict between groups of populations. Sometimes, those could be ethnic riots. Those could be things like political oppression or civil conflicts.

    Then there was actually a third and very extreme category that we looked at, which is really the collapse of civilizations and the breakdown of governing institutions. Now, very few of those events have occurred in modern times, but those scenarios are particularly interesting because they sort of represent extreme scenarios.

    We can think of things like the collapse of the Mayan empire or other empires around the world that coincided with extreme climate changes.


    Professor Solomon Hsiang, thanks for joining us.


    Thanks for having me.