The Carbon Brief, a UK website created in 2011, is a destination for many seeking non-partisan information and analysis on climate change related news and research. I like the neutral tone of the articles and the comprehensive perspective it offers on controversial issues, such as the state of the science on polar bears and, in a similar vein, the growing body of research explaining the "hiatus," a term commonly used to describe a slowdown in the rate of the earth's surface warming. Here's how a Carbon Brief article on one recently published study begins:
This month will see the civil war in Syria reach an inauspicious fourth birthday. Since the uprising in 2011, over 220,000 people have been killed and almost half the Syrian population have fled their homes. Evidence suggests the conflict was triggered by a complex mix of social, political, economic and environmental factors. But new research finds that human-caused climate change could also have had an influence. The study, published in the Proceedings of the National Academy of Sciences, suggests a severe drought that began in 2006 was a catalyst for the conflict, and that climate change has made such droughts in the region more than twice as likely.
In my previous post, I discussed how this new study was broadly covered in the media. I felt that many of the news articles did not critically examine the paper, much less inform readers of the contentious climate/conflict research field it is situated in. The Carbon Brief article, however does a nice job contextualizing the study, while also providing a flavor of the field's conflicting views. Likewise, the UK Climate & Migration Coalition says the new paper adds to a body of contradictory research:
There is currently a raging academic debate about whether a hotter planet will lead to more armed conflict. (We’ve explored this debate in previous blog posts). A number of research papers have found a powerful connection between various climate impacts and armed conflict. Other papers have found a much weaker connection – or no connection at all. Some even found a decrease in violence. Making global generalisations is difficult. However it is a useful exercise to establish whether, in general, climate change will lead to more armed conflict.
How "useful" is the paper? We'll return to that question in a minute. First, I want to point to an article on the study that did feature a critical take and a window into the complexities of the climate/conflict scholarship. Mark Zastrow, reporting in Nature:
The research adds to a fierce debate over whether climate change influences human conflict. In 2009, a study in the Proceedings of the National Academy of Sciences found evidence that warming contributed to the incidence of armed conflict in Africa. A 2013 analysis in Science looked at 60 studies and found that climate change was linked to conflict over a broad span of time periods and geography. But no study has yet described a mechanism by which climate change leads to conflict, a sore point for political scientists who argue that a complex stew of factors drives social unrest.
To understand why all the media attention lavished on recent climate change/conflict research is a sore point for some scientists, read this longish post by Ed Carr, a geographer at the University of South Carolina, where he directs the Humanitarian Response and Development Lab (HURDL). Previously Carr was the climate change coordinator for the Bureau of Democracy, Conflict, and Humanitarian Assistance at the United States Agency for International Development (USAID). He has also been a review editor for the IPCC. In that 2013 post, Carr writes:
So, climate change and conflict is back in the media, seemingly with the strength of science behind it. I’ve been a rather direct, harsh critic of some work on this connection before, at least in part because I am deeply concerned that work on this subject (which remains preliminary) might disproportionately influence policy decisions in unproductive or even problematic directions (i.e. by contributing to the unnecessary militarization of development aid and humanitarian assistance).
Last November, my colleague and friend John Horgan expounded on these concerns--and engaged with inconsistencies in the climate/conflict literature--in several posts at his Scientific American blog. Carr, who we will hear more from momentarily, is still a sharp critic of much published research that suggests climate change has sparked violence and instability in some countries, such as the recent paper on Syria's drought contends.
@EricHolthaus@PaulRobbins15 I have to go with Paul on this. Nearly all climate-conflict work is problematic right now, even in peer review — Ed Carr (@edwardrcarr) March 3, 2015
To quickly recap: This new study that has gotten huge media coverage is careful not to overstate climate change-influenced drought as a factor in the rural upheaval that contributed to Syria's uprising. As the paper's authors write:
Whether it was a primary or substantial factor is impossible to know, but drought can lead to devastating consequences when coupled with preexisting acute vulnerability, caused by poor policies and unsustainable land use practices in Syria's case and perpetuated by the slow and ineffective response of the Assad regime.
At the same time, the authors found that the country's severe drought in the mid-to-late 2000s was "more than twice as likely as a consequence of human interference in the climate system." Based on this, they assert:
For Syria, a country marked by poor governance and unsustainable agricultural and environmental policies, the drought had a catalytic effect, contributing to political unrest.
The Naturearticle by Mark Zastrow stands out from the media pack for the criticism it included of the study:
Some argue that that these failures of policy — not climate change — had the greatest influence over the drought’s impact. Andrew Solow, an environmental statistician at the Woods Hole Oceanographic Institute in Massachusetts, says that the analysis adds little to the debate over the link between climate and conflict. “In the absence of this drought, would there have been violence? And that’s not a question you can answer,” he says. “We don’t have models like that.”
[Lisa Friedman of ClimateWire also interviewed Solow for her story, which is behind a paywall. I included an excerpt from it at the bottom of my previous post] In an email exchange I had this week with Carr, he elaborates on what he believes are the study's strengths and weaknesses:
The attribution of this drought to climate change is tricky, but the climate community is getting a lot better at attribution these days. From the IPCC’s SREX report to now, I would say that there have been interesting improvements in our ability to link the changing probability of particular types of events in particular places to climate change, which in some ways can then speak to the probability that any given event being the product of climate change. But this remains a probabilistic exercise, and we cannot take any one drought/tornado/hurricane and say “climate change!” At least not yet. I don’t think the authors of this study go outside these lines, though I am not an expert on attribution in atmospheric science. Indeed, the attribution exercise seems to be the rigorous part of the article. I say rigorous part because I think the translation of this drought into conflict is pretty weak. Basically, they plumb the conflict literature to support really general statements like "The conflict literature supports the idea that rapid demographic change encourages instability.” No kidding – not sure a citation was needed there. But the causal change between climate change, drought, displacement, and conflict is long and crosses several bodies of data/evidence, all of which are uncertain. The compounding uncertainty in this causal chain is never addressed, so I can’t tell if it is offsetting (that is, some parts of the causal chain address weaknesses in other parts, thereby making the connection throughout the chain stronger) or compounding. I doubt the authors know, either. Basically, I don’t understand how you can get any real understanding of the likely contribution of climate change to this conflict via this mechanism.
I mentioned to Carr that I thought the paper did a good job laying out the socio-political triggers--such as Syrian government failures-- for the country's civil war. He responded:
I think the social stuff in the middle of the paper appears far stronger than it actually is. While there is a lot of interesting, concrete stuff on the Syrian context leading up to the drought, none of it means much of anything because there is nothing here but broad statistics and trends. For example, let’s look at the claim that by "February of 2010, the price of livestock feed had increased by three fourths, and the drought nearly obliterated all herds.” I am willing to accept they have good data on this. But who was impacted by this? Who owned herds? Do rich people or poor people own herds? Rural or urban dwellers? Men or women? In short, this is a severe impact, but on whom? It probably did not have a cross-cutting impact on the entire population. The same could be said for rising food prices. The rich probably managed this fine, but what of the rest? Were there herders and subsistence producers who were able to strategically de-link from markets and weather this fairly easily? Again, who took the hit from this? Even cutting fuel subsidies principally hurts those who own vehicles or those who use them often – so the wealthy and middle class, and principally the urban. Unless they are closely tied to some sort of commodity price impacts, the cutting of subsidies probably had attenuated impacts on the poor in rural areas. Basically, it is just a list of stuff happening in Syria, with the presumption that the confluence of events somehow constituted a trigger for the conflict inside a vague, “Syrian” population. But of course not all of Syria rebelled, and not all that rebelled did so at once. The paper doesn’t even go into the origins of the conflict – who rebelled first and where? How were these first rebels affected by all of these trends? Wouldn’t this be the easiest pathway to identifying the climate-to-conflict connection here? Yet the authors don’t do this at all. It’s all just overgeneralized, semi-circumstantial evidence on the social side.
I wondered if Carr was drilling down too deeply at a granular level. He shot back:
Well, granular is where stressors get translated into responses!
At this point, Daniel Abrahams, who is pursuing his PhD. in the geography department at the University of South Carolina, joined the dialogue. Abrahams works in the Humanitarian Response and Development Lab (which Carr oversees). His research focuses on the interactions between environmental degradation and violent conflict. He chimed in on the Syria drought study's merits:
I agree with the premise that the drought contributed to the ongoing conflict in Syria. What is unknown, and really unaddressed here, is how much. And what can be done about it. Getting good data on the mechanisms that cause conflict is a real challenge for the field. Where climate is most likely to affect conflict (Syria, Yemen, Somalia, etc) are the places where humanitarian and development organizations are the least likely to have access. That is to say nothing of researchers. This presents a bit of a catch 22 for the research community, the places that need to be looked at are the one's we have the least access. This does not mean there aren't other ways to do the research, but we need to be explicit about this shortcoming. As it relates to my research, the people* I have interviewed have been explicit: we understand there is a logical connection between climate change and conflict, we agree with the general themes of the Arab Spring research, but what are we supposed to do about it? Development, humanitarian, and adaptation as sectors are very interested in addressing this issue. But noting simply that drought in Syria affected the conflict does nothing for them operationally.
*Abrahams: "My interviews have been with a range of people either directly involved with, trying to influence, or doing research on policy as it relates to climate change and security — mostly from a US perspective. This includes people working from a defense/intelligence, diplomatic, humanitarian assistance, development, and environment perspective. It is roughly evenly split in terms of those working directly for the USG and non-governmental actors." So where do we go from here? Abrahams completed his thought:
What is needed, really, is a deeper understanding of the social mechanisms that can translate a climate impact into increased likelihood or severity of conflict. And within that, identification of possible inflection points and/or means to build resilience to these events. That requires linking broader climate impacts to a specific biophysical event (which this paper does well) to a social phenomena (migration, resource issues, etc) to a conflict. They glazed over that last step. I understand why from a research perspective, but they need to be way more explicit about why--and that they did. Likewise, I don't think the defense actors I have interviewed could do much with this information either, as they are looking for work that is far more operational. Getting at the social aspects of this work will require a far deeper understanding of a place (ideally multiple places), to understand how drought or other biophysical events can trigger a phenomena and how that phenomena can affect conflict, or for that matter, cooperation. They cannot allude to this side of the work if they didn't explicitly do it. From a policy perspective this really matters. By waving at the social side of things in a non-methodologically sound way, this [paper] calls this work into question, which by proxy calls other' work on this issue into question. Unfair, yes, but clear that this has happened based on my interviews [about similar climate-conflict research published in recent years]. Finally, and what I have been dancing around, is this offers very little to the communities that want to act on on this issue. Yes, we need to know where and how the biophysical events might happen (the contribution of this paper), but we also need to know where to take it from there. I would guess policy makers see this paper as a distraction; something that fills their inbox with people tangentially paying attention to climate issues. We know from a conceptual standpoint how climate change can interact with conflict, what we don't know, and what this paper doesn't answer, is what, if anything, can be done about it.
My takeaways from this exchange with Carr and Abrahams: 1) The social science elements of climate/conflict research lacks rigor; 2) The recently published study linking climate change-related drought to Syria's uprising suffers from this methodological flaw; 3) As a consequence, the environmental security field risks not being taken seriously by policymakers; 4) Getting solid on-the-ground data to understand the climate/conflict relationship is very difficult because of the unstable and dangerous nature of the countries in upheaval; and lastly, 5) It remains to be seen how this research can be applied.
