(Climactic) Conflict of Interest

By Prabarna Ganguly

Rising temperatures.

More heat waves and droughts.

Stronger and more intense hurricanes.

Sea levels rising by 1-4 feet by 2100.

These are the most commonly known negative consequences of climate change.

But did you know that recent research in the field of climate change and policy is arguing for a new link between resource scarcity caused by climate change and increased violent conflicts around the world? In a November 2017 interview with Vox, social psychologist Harald Welzer stated that “we will see a renaissance of violent conflict in the 21st century, and that many of these conflicts will spring from climate change.” While sounding like an alarmist comment, his words are driven by data accumulated by policy makers and scholars that heralds a similar message.

But these data are biased and therefore flawed, says a new paper by Courtland Adams and colleagues, published in February in the peer-reviewed journal Nature Climate Change.

The question is: How come?

Kosher statistical practice predicates that it is integral to sample data in an unbiased manner. For example, an “honest” clinical trial dataset on the effect of treatment X on cancer would include human participants from around the globe. However, this is rarely the case, with most participants usually gathered from the country where the study is done. Similarly, psychological studies should include people of all backgrounds and ages, but undergraduates in university campuses dominate these samples.

Climate change and conflict research is suffering a similar predicament. Many studies published in this field are marred by the “Streetlighting Effect”, which describes how easy access to certain cases or people creates a biased data set.  Researchers have, for instance, focused on the effect of a 2007-10 drought spell on the outbreak of the Syrian war in 2011. However, other countries such as Jordan and Lebanon have been minimally studied, exactly because they did not have such a large-scale violent conflict.

What does such a biased sample mean? The paper explains:

If the evidence of a causal association between climate and violent conflict is informed only by exceptional instances where violent conflict arises and climate also varies in some way, it is unable to explain the vastly more ubiquitous and continuing condition of peace under a changing climate.

Adams and his team are quick to point that the paper does not aim to criticize individual studies (which could have good reasons for focusing on the regions), but the large scale sampling bias does create four major problems. These include:

  1. The impression that the link between climate and conflict is stronger than they actually are.

  2. By focusing only on violent conflict, learning the ways by which equally climate-change affected nations are maintaining and adapting peacefully is being limited.  

  3. Specific places become stigmatized, with African countries such as Kenya and Sudan being most likely to be linked as being more “naturally violent” than others. Trends in violent conflicts must also be studied by including the complex roles played by politics, economics, and other institutions.

  4. Evidence for a positive correlation between climate-change and conflict comes from studying regions that easy to access (such as Sub-Saharan Africa). This diminishes scholarly attention to other afflicted regions, especially in South America and Oceania.   

Assessing how countries are, and will respond to climate-change is an important scope of research. Moving forward, it is going to be critically important to be cognizant of the human bias we include in our methodologies, since they can have long-lasting impact on how nations strategize new ways to tackle the changing environment.  

Read more on WOC!