Tuesday, December 13, 2016

The war over water and how it relates to the catastrophe in Syria


There is a catastrophe in Syria, and, to a lesser extent, Yemen and other parts of the Middle East and sub-Saharan Africa.

Are such catastrophes random? Or are there causal factors which help not only to explain such events, but which also might reduce the chance of similar catastrophes in future?

Below is adapted from a dialogue concerning a paper currently under review, which, in part argues that climate change (via an intensified drought) has contributed to the Syrian civil war. I have anonymised some of it. I am posting it here, publicly, even though it is in such detail. Perhaps someone will have a comment that gives me (or the reader) more understanding and insight.

First, I stress, I am unaware of a single paper suggesting climate change or the Syrian drought is the only factor. Of course, there are pre-existing divisions in Syria. Naturally, these have a historical legacy. But there are grievances and divisions are in any location (Canada, West Bengal etc). A decade ago Syria was comparatively peaceful, even though it had a repressive government then, too; certainly not paradise

I have obscured the names of a couple of people mentioned in the original dialogue.

Dialogue with anonymous reviewer:

critical reviewer: "The second issue is that the two papers (Gleick and Kelley et al) do not make the claim in the aforementioned sentence (which was "It is true that a growing number of commentators from the UK’s Prince Charles onwards, are making a claim, but none of these is based on any analysis of the Syrian conflict, but repeating a now received wisdom" - ie the reviewer claims that neither Gleick nor Kelley et al argue that the drought is linked to conflict). Kelley et al is a straightforward climatological account of drought in the region with a climate science ‘detection and attribution study’ of whether the risk of the 2007-2010 drought was enhanced by underlying climate change. They find that there is a climate signal in that drought event. But at no point in the paper is there any analysis of the causes or impacts of drought on wider society or indeed conflict. (italics added)

The Gleick paper has a wide historical account of the use of water in semi-arid and Mediterranean basin countries including the fertile Crescent region, followed by a description of the 2007-10 drought in Syria, with reference to a few newspaper articles on internal migration. [CB: some of these are more substantial, eg reports by The Center for Climate and Security and another by the FAO].

Reviewer: But that paper again does not make the claim in the commentary that leads from climate change to conflict.

CB: These comments are very interesting and I thank the reviewer for them. First, let me declare that my personal interest in climate change and conflict started almost 3 decades ago (when I read a Lancet editorial in 1989 mentioning this possibility) and from 2010-14 I chaired the supervisory panel of the first PhD thesis, to my knowledge, on the topic of climate change, conflict and public health. I thus have a long and deep knowledge of the relevant literature, have met several key figures and have had relevant articles both accepted and rejected, sometimes with extensive debates with editors and reviewers. I also have corresponded extensively with Dr Peter Gleick, referred to by the reviewer.

The literature on this topic is polarised

My conclusion, based on many years of thought and study, is that the issue of whether climate change is causally related to conflict depends partly on one’s “mental model” of causation and partly on one’s discipline, and the “gatekeepers” to that discipline’s literature. What follows is a long and detailed attempt, by me, to argue that the mental model linking water scarcity to conflict is sufficiently rich to warrant publication in high impact health journals.

Both PNAS (which published the paper by Kelley et al) and Weather, Climate and Society (which published Gleick's paper) may have insisted on very nuanced, cautious language, and I recognise that your journal may also. I have therefore, tried to be less assertive in my revision.

However, Gleick writes, in a section called “The role of climate change stresses”:

“As early as 2008, there were indications that drought frequency and intensity in the eastern Mediterranean area had changed from historical climatic norms. Mathbout and Skaf (2010) used two drought indices [the standard precipitation index (SPI) and the effective drought index (EDI)] to identify an increasing tendency in annual and seasonal drought intensity in all regions corresponding with an increasing number of dry days in the rainy season. Romanou et al. (2010) show statistically significant increases in evaporative water demand in the eastern Mediterranean region between 1988 and 2006, driven by apparent increases in sea surface temperatures. A research paper published in 2012 further suggested that climate change is already beginning to influence droughts in the area by reducing winter rainfall and increasing evapotranspiration (see Figs. 5a,b) (Hoerling et al. 2012). That study suggested that winter droughts are increasingly common and that human-caused climate change was playing a role. Martin Hoerling, one of the study authors, stated, ‘‘The magnitude and frequency of the drying that has occurred is too great to be explained by natural variability alone’’ (NOAA 2013). Whether current conditions are evidence of human induced climate change, the indications for the future water resources of the Middle East and North Africa, including Syria, are not promising.”  (bold italics added)

‘According to the latest IPCC [Intergovernmental Panel on Climate Change] assessment, the climate is predicted to become even hotter and drier in most of the MENA [Middle East North Africa] region. Higher temperatures and reduced precipitation will result in higher frequency and severity of droughts, an effect that is already materializing in the Maghreb.’’

CB: Here, and additionally in this paper (not quoted here) Gleick reviews evidence that climate change is already influencing (for the worse) the frequency of recent, current and future drought. But, he also writes extensively about conflict and water scarcity:

“Water-related conflicts occur in many forms, including disputes over access to water and the control of water systems, the targeting of water infrastructure and systems during conventional conflicts and terrorist actions, and the use of water as a weapon” and “Analysis on the links between possible climatic changes and international security and conflict has been developing for more than two decades, including work on the connections between conflict and stresses from population growth, water scarcity, and agricultural production”.

CB: Gleick, a globally acknowledged expert on water, and a Fellow of the US National Academy of Sciences, is in disagreement with the recently and currently dominant perspective (especially in the political science literature) that conflict does not occur over water; a view, which in my view, is ideological. For example, in late 2014, I attended a lecture by Prof ** (a global leader of the conservative or agnostic view) where he asserted that conflict over water does not occur. At the time ISIS had recently captured control of an important dam, near Mosul in Iraq. This report mentions “The deployment of air power by the US in support of Kurdish forces has shown how seriously the White House takes the potential threat posed by IS control of the dam.”

CB: I asked Prof ** to comment on this apparent counter-example; did it not illustrate an important principle contrary to his general view? Unfortunately, he did not engage with this question.

CB: Threats of war have also been made by senior Pakistani officials against India concerning the Indus Water Treaty for some years (eg "Along the Indus River, saber rattling over water security." Science) including in 2016 and in 2013 by then Egyptian president Mohamed Morsi concerning suggestions to dam the Blue Nile in Ethiopia (eg).

CB: Until the 1980s water competition was considered a trigger for a war involving Israel in 1967 (eg Cooley, J. K. (1984). "(The War Over Water) (a pdf is available here) published in Foreign Policy and cited 209 times according to Google scholar. In parts, it says:

Cooley: “In 1967 Israel went to war against Syria, and Syria's ally, Egypt, partly because the Arabs had unsuccessfully tried to divert into Arab rivers Jordan River headwaters that feed Israel. During that war Israel captured Syria's Baniyas River, the last of the important Jordan headwaters not under Israel's control. Israel also succeeded in destroying the foundations and thus halting construction of a giant new dam at Mukheiba on the Yarmuk River, which runs toward Israel between Jordan and Syria. This dam, still desired by Amman and enjoying American and World Bank (International Bank for Reconstruction and Development) backing, would have greatly augmented the water available to pre-1967 Jordan, including the West Bank, but might have deprived Israel of water supplies its planners coveted...”

CB: The view that conflict over water is no longer important has in recent years changed. Another widely cited author, Wolf (2007), wrote (in: "Shared waters: conflict and cooperation." Annual Review of Environment and Resources 32: 241-269.)
Wolf: "Although wars over water have not occurred, there is ample evidence showing that water issues have led to intense political instability and that acute violence has occasionally been the result. Conflicts over shared water resources occur at multiple scales, from sets of individual irrigators, to urban versus rural uses, to nations that straddle international waterways."

"Within this framework, water resources lead to intense political pressures, while threatening the processes of sustainable development and environmental protection."

CB: While Wolf’s paper suggests sympathy for Gleick’s argument his paper referred to above was cited in the most recent IPCC report (in the chapter on Human Security) in support of a more conservative view:

"shows strong evidence of significant formal cooperation among river basin riparian states..".

CB: I find this IPCC conclusion (reflected by this extract, and consistent with dominant political science discourse) as ideological, as if to acknowledge the risk of conflict over water might legitimise or perhaps justify war. But there is a counter-view, which I also subscribe to: to suppress the causal link between water scarcity, competition and possible conflict, may instead foster complacency which actually heightens the risk of future conflict. One could argue that the position I hold, which (likely) Prince Charles also does is similarly “ideological”, but one could also argue that each view (i.e. resource scarcity drives co-operation or resource scarcity drives conflict) is valid. I actually believe scarcity can drive co-operation, but beyond a threshold, conflict is more likely (one reason I tried to introduce the concept of thresholds in the original version of the paper). I believe competition over limited resources is fundamental to the theory of evolution, and also consistent with “common sense” (imagine 15 strangers arrive at your house for dinner tonight, or, if that does not disturb you, imagine 150 strangers who are very hungry.)

CB: Gleick reviews examples of competition over water and conflict dating to 4500 years ago, in a section called “Historical background: Mesopotamia, water, and conflict”

“In the current civil war, some analysts have argued that factors related to drought, including agricultural failure, water shortages, and water mismanagement, have played an important role in contributing to the deterioration of social structures and spurring violence (Femia and Werrell 2013; FAO 2012; Mhanna 2013). In particular, the combination of very severe drought, persistent multiyear crop failures, and the related economic deterioration led to very significant dislocation and migration of rural communities to the cities. These factors further contributed to urban unemployment and economic dislocations and social unrest.” (bold italics added)

CB: Gleick goes into even more detail, which I will not fully describe here. For example:

”That cable (by the US embassy in Syria in 2008) describes a briefing by FAO Syrian Representative Abdullah bin Yehia on drought impacts, which he described as a ‘‘perfect storm’’ when combined with other economic and social pressure. Concerns expressed at that time also noted that the population displacements ‘‘could act as a multiplier on social and economic pressures already at play and undermine stability in Syria.’

Gleick: “some of the earliest political unrest began around the town of Dara’a, which saw a particularly large influx of farmers and young unemployed men displaced off their lands by crop failures.” 

Gleick: “It is logical to conclude that escalating pressures on urban areas due to internal migration, increasing food insecurity, and resultant high rates of unemployment have spurred many Syrians to make their political grievances publicly known.” … In all three centers of popular uprisings lie important narratives of livelihoods lost and families left wanting’’ (Saleeby 2012).

He concludes:

Gleick: "The Syrian conflict that began in 2012 has many roots, including long-standing political, religious, and social ideological disputes; economic dislocations from both global and regional factors; and worsening environmental conditions."

CB: This paper thus argues that key environmental factors for conflict include both direct and indirect consequences of water shortages, ineffective watershed management, and the impacts of climate variability and change on regional hydrology.

CB: The reviewer disputes that (Gleick (and Kelley et al) would support the claim that “a growing number of authors have argued that climate change aggravated the Syrian drought which in turn helped to precipitate the bitter civil war”.

CB: However, I disagree with Reviewer 2’s interpretation of Gleick’s paper, which in my view (as I have tried to show) clearly links climate change to the worsening drought and in turn links that to conflict.

CB: Gleick also wrote a paper about this topic in the non-peer-reviewed literature (Gleick, P. (2014). "Drought, Water and Agricultural Management, and Climatic Conditions as Factors in the Syrian Conflict." Huffington Post.) In this, he wrote, in part:

Gleick: “The drastic decrease in water availability, water mismanagement, agricultural failures, and related economic deterioration contributed to Syria's population dislocations and the migration of rural communities to nearby cities. These factors further contributed to urban unemployment, economic dislocations, food insecurity for more than a million people, and subsequent social unrest.” (emphasis added).

CB: Kelley and three co-authors in the PNAS (2015) paper cover some similar territory to Gleick. Contrary to the opinion of the reviewer, they also link the recent drought in Syria not only to climate change, but also to conflict, for example, in their abstract they conclude:

Kelley et al:
“Analyses of observations and model simulations indicate that a drought of the severity and duration of the recent Syrian drought, which is implicated in the current conflict, has become more than twice as likely as a consequence of human interference in the climate system.”

CB: Social science is imprecise; however, the words “implicated in the current conflict” can be interpreted to suggest causality, analogous to legal reasoning e.g. “Lee Harvey Oswald was implicated in the murder of John F Kennedy”. It is not “proof” of causality in a Bradford Hill sense, but it is suggestive. Furthermore, even in epidemiology, “proof” is often elusive. Additional forms of evidence might help convince a sceptic of the causal relationship between the Syrian drought and conflict, but these are difficult to obtain, and a richer description of the putative causal framework is beyond the scope of my paper. However, I argue, the link is plausible and important. 

CB: Evidence that there is no causal link is even weaker than the converse. And, again, I dispute the reviewer’s assertion that Kelley et al would not support the claim that “a growing number of authors have argued that climate change aggravated the Syrian drought which in turn helped to precipitate the bitter civil war”.

CB: Kelley et al's paper has a summary box. In part it reads:

“There is evidence that the 2007−2010 drought contributed to the conflict in Syria. It was the worst drought in the instrumental record, causing widespread crop failure and a mass migration of farming families to urban centers. Century-long observed trends in precipitation, temperature, and sea-level pressure, supported by climate model results, strongly suggest that anthropogenic forcing has increased the probability of severe and persistent droughts in this region, and made the occurrence of a 3-year drought as severe as that of 2007−2010 2 to 3 times more likely than by natural variability alone. We conclude that human influences on the climate system are implicated in the current Syrian conflict."

Statistical papers linking climate change and conflict - take care

Kelley et al:
"Our thesis that drought contributed to the conflict in Syria draws support from recent literature establishing a statistical link between climate and conflict (25–27).

CB: I am aware of this statistical work, but have not cited it in my paper, not only due to lack of space, but also because I am aware of methodological criticisms of some of these papers that seem plausible to me (i.e. here I am less supportive than Kelley et al). 

CB: However, the scarcity of convincing statistical studies that so far link climate change and conflict is not, for me, a major obstacle. Certainly any studies which purport to statistically show the reverse (i.e. no link) are at least as problematic. Kelley et al comment:

Kelley et al:
“A more fundamental objection (27) is that data-driven methods do not provide the causal narrative needed to anoint a “theory” of civil conflict, and the quantitative work on climate and conflict has thus far not adequately accounted for the effects of poor governance, poverty, and other sociopolitical factors.”

CB: This is probably true. That is, quantitative work on the topic of climate and conflict is problematic, not least as climate change may either improve or harm resource availability. But this does not negate the possibility that adverse climate change may exacerbate resource scarcity, and in some cases fuel conflict. That pathway is, to me, highly plausible on theoretical grounds, and also empirically, where climate change reduces resource availability. Were climate change to increase resource availability (e.g. crop yield in some favourable locations with suitable soil) then it could plausibly reduce conflict, though again this likelihood must be quantified, for example such a benefit could be negated by poor governance, or if more resources attract in-migration which causes tension.

CB: Finally, though it may not convince the reviewer, Kelley et al recount:

"in a recent interview (45), a displaced Syrian farmer was asked if this was about the drought, and she replied, “Of course. The drought and unemployment were important in pushing people toward revolution. When the drought happened, we could handle it for two years, and then we said, ‘It’s enough.’”

CB: This is common sense, albeit “only” an anecdote. But if the view is commonly held, then it is plausible it points to a causal factor. We need strong evidence to support the alternative hypothesis – such evidence is lacking.

In summary, thank you for reading this carefully. I believe I have presented a very thorough case for publication and I believe this is an extremely important issue for global and planetary health. Although I believe the revised paper is now clear, if succinct, I would of course be willing to undertake additional revision.

This debate matters a great deal. If the link between climate change and conflict is denied then it would follow that we need not be very concerned .. simply spread air conditioning (for heatwaves) or insecticides (for vector borne diseases). Actually, it is ridiculous to dispute the link between climate change, drought and conflict, at least where regional limits to growth are close.