The publication of the report from Working Group II of the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) on 31 March 2014 was accompanied by considerable media publicity, some it suggesting that the world was facing an era of violent upheaval. However, the main discussion of conflict, which is found in the chapter on human security, is moderate in tone and cautious in its conclusions. Other chapters in the WG II report use more dramatic language, while a methods chapter completely dismisses the link between climate change and conflict.

The day after the publication of the most recent report from the UN’s Intergovernmental Panel on Climate Change (IPCC) on the effects of climate change, the Norwegian daily newspaper Dagsavisen was able to report that Norway’s Minister for Climate and Environment now envisaged a future world with more conflicts.Footnote 1 This is in line with claims made earlier by the Norwegian Nobel Committee. Against this background, I embarked with some anticipation on the report’s 2,679 pages. I found that each of the four chapters that address this question gives a slightly different answer.

The question is discussed most thoroughly in the chapter on human security. This IPCC report is the first to contain such a chapter. The report defines human security broadly, in my opinion far too broadly, but a separate sub-section is devoted to violent conflict. This latter subject was mentioned sparsely in the previous two IPCC reports (published in 2001 and 2007), to some extent on the basis of weak sources. This time the scope of the sources is wider, but at the same time more stringent. The chapter concludes that while some studies associate warming and variable precipitation with violent conflict, other studies do not. Accordingly, there is no basis overall for one to conclude that there is a strong connection. This view is consistent with previous summaries of the literature.

The chapter also points out that climate change is generally believed to influence a number of factors that are frequently associated with violent conflict, such as poverty, poor economic growth and misgovernment. This theme recurs in several other chapters. As long as nothing more precise can be said about the strength of the connection in each of the two causal stages (from climate change to risk factors, and from risk factors to violent conflict), we cannot conclude with any certainty about the role of climate change in violent conflict.

In the 2007 report, the chapter on Africa was the source for the most dramatic assertions about possible connections between climate and violent conflict. The tone is more cautious this time. Although there are several references to the possibility of an increase in violent conflict, the report also highlights the disagreement that exists between researchers.

A methodological chapter, under the somewhat dry heading ‘Detection and attribution of observed impacts’, gives an extremely critical assessment. The chapter builds on all the others, and evaluates whether the available material provides a basis for robust conclusions about a connection between climate change and its presumed effects. The authors point out that all phenomena that are believed to be influenced by climate change are also influenced by many other factors for which it may be difficult to control.

When the authors come to the question of conflict, little is left standing. First, there is a pervasive uncertainty in the literature regarding the empirical findings. And secondly, most empirical research has focused on annual variations in temperature and precipitation instead of deviations from long-term averages. Thus, these studies are more about variations in the weather than about climate change. Hence, this chapter concludes that we cannot say anything certain about the existence or magnitude of climate-change effects on violent conflict. This conclusion is repeated in its entirety three times: first for civil war, then for small-scale communal violence, and finally for violent individual crime.

A chapter on ‘emergent risks’ is more alarmist. This chapter presents climate change as a potentially significant factor for future conflict. But once again there is an emphasis on the fact that there is ‘low confidence’ as to the existence of any documented effect of climate change—as oppose to climate variations—on conflict.

I was anxious to see what use the report would make of a controversial article that appeared in Science last autumn. That article, which was published just before the IPCC’s literature cut-off date, asserted in broad terms that climate in general, and warming in particular, was a significant factor in conflict at all levels, from individual aggression during a heat wave to international warfare and regime collapse on a millennial scale. The article claimed to be the first ‘meta-study’ of the field, or in other words, the first comparative statistical assessment of results from all relevant previous studies. Most of the authors whose studies were summarized, however, found it hard to recognize the presentation of their own work. A collective response from 26 researchers (including the author of this article) is now in press. But since the response will not appear in print until several months after the IPCC’s literature cut-off date, it is obviously not referred to. Nevertheless, the authors of the chapter on human security would have been aware of the debate—among other things, the article was heavily criticized by several leading German climate researchers in a wide-ranging report published in Der Spiegel on 1 August last year. In the human security chapter, the controversial Science article is treated as one of several summaries of the literature. By contrast, the chapter on ‘emergent risks’ (which is co-authored by one of the three authors of the Science article) presents the article as a contribution with a higher status.

Two summary chapters, the Technical Summary (TS) and the Summary for Policymakers (SPM), contain formulations that are close to those used by the chapter on human security. Regarding conflict, the TS refers to several of the other chapters, but not to the methodological chapter. There is also a significant linguistic nuance in that the TS claims that climate change will increase risks of violent conflict, while the SPM claims that climate change can increase such risks. The use of word such as ‘can’ or ‘may’ in academic writing is extremely problematic, as it provides no basis for evaluating the probability of an event, beyond the fact that it is not zero.

The TS and the SPM are no doubt the most politically influential parts of the report. Shortly before the report was made public, it emerged that one of the two coordinating lead authors of the chapter on economic effects, Richard Tol, had withdrawn from further work on the SPM because he thought that the summary articulated a pessimism for which there was no basis in the individual chapters.

The IPCC’s view as to the risks of climate change leading to violent conflict thus depends to some extent on the chapter one chooses to rely on. In my opinion, the methods chapter is the most solidly based, but like the authors of last year’s controversial Science article, I am not an impartial observer. In any event we can be confident in saying that the IPCC report does not put forward a consensus that climate change will lead to more wars.