US economists have drawn together 45 sets of evidence spanning 10,000 years to show that warmer temperatures and more extreme rainfall can cause greater human conflict. University of California, Berkeley’s Ted Miguel says this “could have critical implications for understanding the impact of future climate change on human societies”.
“Many global climate models project global temperature increases of at least 2°C over the next half century,” Ted told me. “Our findings suggest that global temperature rise of 2°C could increase the rate of intergroup conflicts, such as civil wars, by over 50% in many parts of the world, especially in tropical regions where such conflicts are most common.”
Scientists have long puzzled over whether data backs climate as a cause of violent events such as the fall of the Roman empire. Global warming has brought an ‘explosion’ of interest from researchers, from archaeologists to psychologists, in climate-linked violence. And the types of conflict vary from fights between two people to civil wars and collapse of whole civilisations. But some studies see political, economic and geographical factors as more important than climate.
How researchers assess their data in these studies could introduce problems. For example, it can often be argued that ‘correlation does not imply causation’, meaning that links between two data sets might be caused by other factors. For example, reading ability might seem to improve as shoe size does, but one doesn’t cause the other – getting older causes both. To find any real, bizarre, link between shoe size and reading ability, you would need to look at people with the same age – or ‘control for’ age.
So Ted and his Berkeley teammates only brought together data that could be used to find causal links, although not all the original studies they started from had done this. Bringing together data from many different studies, collected all over the world, considering different types of violence, gives their findings stronger backing than each lone study. They called on records collected in many places that had taken measurements repeatedly in each place, and analysed them from scratch to reach their own conclusions.
“This data allows for one to account for fixed factors that differ across locations and time,” Ted explains. “We end up just relying on variation that is much closer to a ‘natural experiment’, for instance, by comparing crime rates in New York City in a year where the month of July is anomalously hot versus a month of July that is unusually cool. This type of data makes it possible to use the latest, most credible, research methods in the analysis. These tools are becoming ever more widely used in social science research, but were not universally employed in the studies we reanalysed, especially in some of the older studies and for those in fields where these tools are only now becoming popular.”
The economists also deliberately don’t control for factors like average income, which can themselves be affected by climate. “In many cases this approach is more harmful than helpful,” they write in a paper just published in top research journal Science. “At the extreme, if temperature influences conflict only through income, then controlling for income would lead the researcher in this example to draw exactly the wrong conclusion about the relationship between temperature and conflict: that there is no effect of temperature on conflict.”
From data covering the period from 8000 BC to the present day, Ted’s team finds climate change has a ‘substantial’ influence on violence. They see global patterns of conflict linked to climatic changes like increased drought or higher than average annual temperature. Spikes in domestic violence are repeated in India and Australia, as are ethnic violence in Europe and South Asia. Assaults and murders increase in the United States and Tanzania, as do civil conflicts throughout the tropics. Even the collapses of the Mayan and Chinese empires are linked to climate. “There is currently much debate about the nature of this link, and our results provide the most comprehensive evidence yet assembled on this critical question,” Ted said.
All 27 modern data sets indicate that warmer conditions generate more conflict, a result that would be extremely unlikely to occur by chance alone if temperature had no effect on conflict. But to find out the size of the effect, the economists had to take into account that some places normally have wider temperature and rainfall ranges than others. To do that they compared changes against a yardstick: the statistical measure of standard deviation, which shows how much temperature and rainfall data spread out from a location’s average.
Ted’s team found that for each standard deviation change toward warmer temperatures or more extreme rainfall, median average estimates indicate that the frequency of interpersonal violence rises 4%. The frequency of larger scale conflict rises 14% for each standard deviation. Even worse, the places where we live are expected to warm by 2-4 standard deviations by 2050. “Amplified rates of human conflict could represent a large and critical impact of anthropogenic climate change,” the economists write.
That does assume that we will continue to respond to climate change in the same way we have in the past. As we don’t know yet how this effect works, that’s impossible to say. But ignoring it for that reason would be “a dangerously misguided interpretation of the available evidence”, they write. In the meantime the researchers say that understanding how temperature might increase violence – for example by tilting our mood to the aggressive, or crippling farming incomes – is especially important.
University of Colorado’s John O’Loughlin, who previously found climate’s link to violence in Africa was weaker than other factors, calls Ted’s team’s effort to bring these studies together ‘heroic’. “They certainly have raised the bar for quantitative studies of climate and conflict and taken a strong stance that other researchers will be forced to address,” he told me. “But I fear that they are overemphasising the climate effects on conflict in the face of a lot of other evidence that local social conditions are the dominant set of factors. That’s not to say the climate effect is insignificant.”
That other factors might play a part is not lost on Ted’s team. “We do not conclude that climate is the sole – or even primary – driving force in conflict,” they write. “But we do find that when large climate variations occur, they can have substantial effects on the incidence of conflict across a variety of contexts.” And the world should take note of those likely effects, Ted said. “Taken together, the findings indicate that the risk of future conflict and violence is another reason why serious efforts must be made now to deal with future climate change.”
Solomon M. Hsiang, Marshall Burke, Edward Miguel (2013). Quantifying the Influence of Climate on Human Conflict Science DOI: 10.1126/science.1235367