- Armament and disarmament
- Conflict and peace
- Peace and development
A recently published report by the International Panel on Climate Change (IPCC) is the first to address the question of whether climate change is a causal factor for violent conflict. While earlier IPCC assessments only touched on this issue briefly—for instance by indicating the risk of future water wars—a number of sections of the new report deal with the consequences of climate change for the incidence of violence. However, not all sections of the new report give the same message.
The messages communicated in the various chapters of the Working Group II (WG II) contribution to the IPCC Fifth Assessment Report, Climate Change 2014: Impacts, Adaptation, and Vulnerability, are subtly different. These differences are important because some support the view that climate change will be a driver for more conflict in the future, while others explicitly reject such claims, stressing the lack of knowledge about the consequences of climate change for conflict.
There are references to conflict in many chapters of the WG II report, although these tend to be vague and only weakly supportive of a future link between climate change and conflict. However, two chapters are particularly relevant here: chapter 12 on ‘Human security’ (PDF) and chapter 19 on ‘Emergent risks and key vulnerabilities’ (PDF).
Chapter 12 contains the longest and most detailed treatment of the literature on violent conflict. A large number of studies are referenced and discussed. Some of these studies find a link between indicators of climate change and various types of conflicts, while others either fail to find such a link or are inconclusive. Therefore, the chapter authors’ conclusion is not very surprising:
"… collectively, the research does not conclude that there is a strong positive relationship between warming and armed conflict."
The authors of chapter 19, having devoted less attention to conflict and quoting fewer references, come to a different conclusion. One study in particular is highlighted—the first (and so far only) meta-analysis of the link between climate change and violent conflict, by Solomon Hsiang, Marshall Burke and Edward Miguel. According to the authors of chapter 19, this study ‘implicates climatic events as a contributing factor to the onset or intensification of several types of personal violence, group conflict and social instability in contexts around the world.
Largely based on Hsiang, Burke and Miguel’s study, the chapter concludes that ‘there is potential ceteris paribus [all other things being equal] for large relative changes to global patters of personal violence, group conflict and social instability in the future’ and that ‘the effect of climate change on conflict and insecurity has the potential to become a key risk’.
Several reasons could explain this difference between the two chapters. One is that there may be a difference between assessing knowledge based on current data and expectations of likely effects in the future. Climate change has been minor so far compared to what is expected to come. However, if there is solid evidence for an effect of climate change on violent conflict, as purported in chapter 19, this should also have been the conclusion in chapter 12. The authors of chapter 19 could have than built on this and made their risk assessment.
Alternatively, the authors of chapter 12 are correct, in which case it would have been difficult to come to chapter 19’s conclusions. Another reason could be a lack of communication between the authors of the two chapters, leading one group to disregard the reception of the state of research by the other. However, the study by Hsiang, Burke and Miguel is also discussed in chapter 12.
A more likely explanation is that the authors of chapter 12 and 19 differed in their assessment of what is currently an intense academic controversy. The controversy itself is enacted at conferences as well as in the pages of leading academic journals such as the Proceedings of the National Academy of Sciences (PNAS), Science, Nature and Climatic Change.
Roughly speaking, there are currently two main camps in this debate: ‘ecologists’ and ‘conflict researchers’. The ecologists are certain that the link between violence indicators and climate change has been established beyond reasonable doubt. For example, Hsiang, Burke and Miguel are convinced that they have found ‘a marked convergence among quantitative findings from a broad range of disciplines’. They agree that many questions remain. But these no longer concern whether climate change is a driver of conflict but instead how it does so. In their study, published in Science, they write:
[E]xisting research has successfully established a causal relationship between climate and conflict but is unable to fully explain the mechanisms.
Many of their critics (including myself) have come to climate change science having worked as conflict researchers, and are surprised by their boldness. There are two main critiques of the conclusions of Hsiang, Burke and Miguel. The first has to do with a number of technical choices—including statistical models, selection of data, and other methodological issues—on which the ecologist camp’s conclusions rest (and it should be noted that while the Science study is the focus of debate, there are other controversies).
Conflict researchers have demonstrated that with a few changes in modelling, the strong positive result attained by Hsiang, Burke and Miguel disappears. The conclusion that results depend on model specification and data choices is unavoidable.
The second reason is more philosophical. While ecologists acknowledge that climate change is not the only driver of conflict and that fuller models are needed, they prefer to exclude other variables in testing for the link between climate change and conflict. Consequently, in a modern form of environmental determinism, they have no qualms about make predictions of future conflict on the basis of expected climate change.
In contrast, conflict researchers do not deny the possibility that climate change may drive conflict, but argue that the link has not been sufficiently established in proper conflict models. In technical terms, they are sceptical of relying on models in which climate change or climate variability is the only independent variable (besides fixed effects), preferring models which include control variables that have previously been found to have an effect on conflict.
Fundamentally, conflict researchers recommend extending established conflict models to include climate variables rather than to search for what are, in essence, bivariate correlations between in-unit variances in climate (and often weather) and conflict.
In addition to the detailed technical chapters, IPCC reports contain a summary for policymakers. The authors of the summary of the WG II report, who may or may not have been aware of the differences between the two chapters in the report on conflict, have chosen a problematic middle path. In two sentences, they acknowledge the study by Hsiang, Burke and Miguel (in the second sentence) and refrain from following its conclusion of a causal relationship. Instead they stress drivers prominent in conflict research and, crucially, use the non-committal word ‘can’ in giving their assessment:
Climate change can indirectly increase risks of violent conflicts in the form of civil war and inter-group violence by amplifying well-documented drivers of these conflicts such as poverty and economic shocks (medium confidence). Multiple lines of evidence relate climate variability to these forms of conflict.
This last sentence indicates where research on the links between climate change and violent conflict has to go. Focusing on pathways between climate change and violent conflict will hopefully clarify some of the discrepancies currently dividing the field.
This blog post is published as part of a collaborative partnership between SIPRI and Economists for Peace and Security (EPS).