An integrated model of climate and environmental pathways to violent extremism
Climate change does not turn people into extremists by itself. With Nora Benningstad, Milan Obaidi and Gulnaz Anjum, I reviewed 45 studies and found that it works as a threat multiplier: heat, droughts, floods and stories of ecological collapse feed violent extremism through material hardship, failing institutions, ideological exploitation and psychological threat, and mainly where resilience and governance are weak.
The paper, in plain language
Past a yes-or-no question
Whether climate change causes violent extremism is the wrong question, or at least an incomplete one. The research on it is scattered across economics, political science, terrorism studies and psychology, and the parts rarely speak to each other. We ran a preregistered systematic review following PRISMA guidelines. Of 7,362 records, 45 studies in 40 publications met our criteria: qualitative analyses, event and survey data, and ten experiments. The studies were too different to pool in a meta-analysis, so we used them to build a model of how the pathways connect.
Four routes in
The evidence points to four routes. Material stress: hotter and drier spells are linked to more terrorist attacks and fatalities in several settings, rainfall shortfalls to attacks in Nigeria and Somalia and to religious riots in India, but only in rain-fed areas without dams. Political-institutional stress: after cyclones, typhoons and floods, insurgent groups move into the space the state leaves, and after Pakistan's 2010 floods votes shifted towards a Taliban-linked party where government relief was poorest. Ideological instrumentalism: eco-fascist movements in the Nordic countries and white-supremacist online spaces recast environmental decline as a story about nation and race. Psychological stress: in experiments in Germany, the UK, the US, Austria and Argentina, reminders of the climate threat increased authoritarian aggression, racism and ethnocentrism.
Where the evidence is thin
We sort outcomes along a ladder from endorsing violence through advocating it and being willing to use it to carrying it out. Climate variability and shocks show up mostly at the top rung, in violent acts; environmental change and threat cues show up mostly lower down, in attitudes and propaganda. No study we found measured willingness at all, and none linked climate threat cues directly to recruitment or attacks. The geography is lopsided: studies of weather and disasters come largely from the Global South, while studies of narratives and threat perceptions are almost all from Western countries. Two thirds of the event-based studies lump all ideologies together, the search covered English-language work only, and a call for unpublished studies turned up nothing.
What follows for prevention
Nothing here makes extremism inevitable; most people who live through a drought or a flood never come near it. What the model does is name the conditions under which the risk grows, among them weak household resilience, dependence on rain-fed farming, poor or unfair disaster response, ethnic discrimination and authoritarian predispositions. It also yields testable predictions, such as that the link between disasters and violence should weaken where disaster governance is strong. That makes fair aid distribution, livelihood support and community resilience prevention tools in their own right, and it suggests that inclusive stories of what people can do together are safer than apocalyptic framing.