Climate change and the Syrian civil war revisited
Jan Selby a, *
, Omar S. Dahi b
, Christiane Fr€ohlich c, d
, Mike Hulme e
a
Department of International Relations, University of Sussex, Brighton, BN1 9QN, UK
b
School of Critical Social Inquiry, Hampshire College, 893 West Street, Amherst, MA 01002, USA
c
Center for Earth System Research and Sustainability (CEN), University of Hamburg, Grindelberg 7-9, 20144 Hamburg, Germany
d
Institute for Peace Research and Security Policy (IFSH), University of Hamburg, Beim Schlump 83, 20144 Hamburg, Germany
e
Department of Geography, King's College London, Strand, London, WC2R 2LS, UK
a r t i c l e i n f o
Article history:
Received 26 September 2016
Received in revised form
20 March 2017
Accepted 24 May 2017
Keywords:
Climate change
Syria
Drought
Civil war
a b s t r a c t
For proponents of the view that anthropogenic climate change will become a ‘threat multiplier’ for
instability in the decades ahead, the Syrian civil war has become a recurring reference point, providing
apparently compelling evidence that such conflict effects are already with us. According to this view,
human-induced climatic change was a contributory factor in the extreme drought experienced within
Syria prior to its civil war; this drought in turn led to large-scale migration; and this migration in turn
exacerbated the socio-economic stresses that underpinned Syria's descent into war. This article provides
a systematic interrogation of these claims, and finds little merit to them. Amongst other things it shows
that there is no clear and reliable evidence that anthropogenic climate change was a factor in Syria's precivil
war drought; that this drought did not cause anywhere near the scale of migration that is often
alleged; and that there exists no solid evidence that drought migration pressures in Syria contributed to
civil war onset. The Syria case, the article finds, does not support ‘threat multiplier’ views of the impacts
of climate change; to the contrary, we conclude, policymakers, commentators and scholars alike should
exercise far greater caution when drawing such linkages or when securitising climate change.
© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
1. Introduction
In the view of many Western policymakers and commentators,
the Syrian civil war was caused, in part, by anthropogenic climate
change. Former US President Barack Obama claimed that climate
change-related drought ‘helped fuel the early unrest in Syria, which
descended into civil war’ (Obama, 2015); former Secretary of State
John Kerry argued that ‘it's not a coincidence that immediately
prior to the civil war in Syria, the country experienced its worst
drought on record’ (Kerry, 2015); erstwhile Democratic presidential
candidates Martin O'Malley and Bernie Sanders have claimed
similarly (Democracy Now!, 2015; Schulman, 2015); and in the UK,
Prince Charles has maintained that ‘there is very good evidence
indeed that one of the major reasons for this horror in Syria was a
drought that lasted for five or six years’ (Mills, 2015). International
organisations (e.g. the World Bank: Verme et al., 2016: p. 33),
leading NGOs (e.g. Friends of the Earth: Bennett, 2015), official
governmental and intergovernmental reports (e.g. Adelphi et al.,
2015; King et al., 2015), defence think tanks (e.g. CNA Military
Advisory Board, 2014: pp. 13e14), academics (e.g. Cole, 2015;
Malm, 2016), activists (e.g. Brand, 2015) and commentators of
various political persuasions (e.g. Box & Klein, 2015; Friedman,
2012, 2013) e all have argued similarly.
For its advocates, this Syria-climate change thesis is powerful
not so much for its own sake, but because it illustrates the chaos
that may ensue as greenhouse gas emissions rise. Climate change,
runs the common policy refrain, is a ‘threat multiplier’ (CNA
Military Advisory Board, 2007: p. 44) which will cause ‘more
drought, more famine, more mass displacement e all of which will
fuel more conflict for decades’ (Obama, 2009). The Syria case appears
to confirm this, showing that the conflict effects of climate
change are already with us, and lending extra credibility to warnings
of future climate-driven instability. The Syria example, in turn,
has potentially important policy implications, especially for the
ways in which political, military and development institutions
might prepare for and adapt to the changing global climate. The
Syria-climate change link has been widely invoked, for example, in
* Corresponding author.
E-mail addresses: j.selby@sussex.ac.uk (J. Selby), odahi@hampshire.edu
(O.S. Dahi), christiane.froehlich@uni-hamburg.de (C. Fr€ohlich), Mike.hulme@kcl.ac.
uk (M. Hulme).
Contents lists available at ScienceDirect
Political Geography
journal homepage: www.elsevier.com/locate/polgeo
http://dx.doi.org/10.1016/j.polgeo.2017.05.007
0962-6298/© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Political Geography 60 (2017) 232e244
discussions about Europe's migrant and refugee crisis, with European
Commission President Jean-Claude Juncker (2015) identifying
climate change as one of the ‘root causes’ of the new migration,
others suggesting that those displaced Syrians arriving in Europe
are ‘climate migrants’ and ‘climate refugees’ (e.g. Baker, 2015;
Dinshaw, 2015b), and still others arguing that the numbers
currently arriving in Europe will inevitably rise as the planet warms
(e.g. Hockenos, 2015; O'Hagan, 2015).
For all this, there is good reason for caution about the Syriaclimate
change thesis. Until a few years ago, the 2003e05 war in
Darfur was widely identified by Western commentators and policymakers
as climate change-related e and even as the ‘first climate
war’ (e.g. Mazo, 2010: pp. 73e86; Welzer, 2012: pp. 61e5) e with
UN Secretary General Ban Ki Moon going so far as to claim that ‘the
Darfur conflict began as an ecological crisis, arising in part from
climate change’ (Ki Moon, 2007). But such claims have since been
discredited, with critics finding among other things that Darfur's
war neither occurred during nor was directly preceded by drought
(Kevane & Gray, 2008); that there existed no solid evidence linking
the Sahelian drought to anthropogenic climate change, in fact
possibly the opposite (Dong & Sutton, 2015); and that claims like
those of the UN Secretary General misrepresented the political and
economic causes, and the essentially counter-insurgency character,
of the Darfur war (Verhoeven, 2011; Selby and Hoffmann, 2014a).
More broadly, there is no consensus within the growing field of
climate-conflict studies on whether violence and civil conflict are in
any way related to climatic variables. Although some studies have
identified such linkages (e.g. Hendrix & Salehyan, 2012; Hsiang
et al., 2011), others have concluded to the contrary (e.g. Buhaug,
2010; Theisen et al., 2011/12) e as scientific reviews of the field
have repeatedly shown (see esp. Field et al., 2014: ch. 12; Gleditsch
& Nordås, 2014; also Selby, 2014). Historically, moreover, public and
policy discourse on the security and geopolitical implications of
climate change has been well ahead of, and often at variance with,
the available scientific evidence (Nordås & Gleditsch, 2007; Selby
and Hoffmann, 2014b). Given this background, it cannot be just
assumed that the Syria-climate conflict story is valid: further critical
scrutiny is required.
This paper seeks to provide this, offering a systematic interrogation
of the claimed links between anthropogenic climate change
and the Syrian civil war. We start by summarising the evidence for
this thesis before turning, in the main body of the article, to the
three principal sub-theses which inform it: that anthropogenic
emissions of greenhouse gases contributed to Syria's drought; that
this drought led to large-scale migration; and that this droughtrelated
migration was an important factor in Syria's early unrest.
Each of these sub-theses we find to be seriously flawed; and on this
basis we conclude that there exists no convincing evidence that
climate change contributed to the onset of Syria's civil war and that,
unless new evidence emerges, the Syria case does not support
‘threat multiplier’ views of the impacts of climate change.
2. The Syria-climate conflict thesis
Three separate studies provide the underpinning evidence for
the Syria-climate conflict thesis: (1) a two page briefing document
by Francesco Femia and Caitlin Werrell of the Center for Climate
and Security in Washington DC (Femia & Werrell, 2012) which,
despite its brevity, was the primary reference point for proponents
of the thesis up until 2015 (see e.g. Friedman, 2012; Quinn & Roche,
2014); this briefing has subsequently been extended and published
in peer-reviewed form (Werrell et al., 2015); (2) a peer-reviewed
article by Peter Gleick, one of the foremost scholars of water issues
worldwide (Gleick, 2014); and (3) a further peer-reviewed
article by Colin Kelley and colleagues, mostly earth scientists at
the universities of California and Columbia (Kelley et al., 2015). All
three studies cover similar territory, and turn to many of the same
sources. The latter study stands out, however, in having been
published in the Proceedings of the National Academy of Sciences,
and in being the only one of the three studies to deploy climate
modelling. Its finding that climate change is ‘implicated’ in the
onset of Syria's civil war was immediately and extensively reported,
in outlets ranging from Nature (Zastrow, 2015) to the Daily Mail
(Gray, 2015) e indeed it is already one of the top ten most mediacited
climate change studies of recent years (McSweeny, 2015) e
such that it has now become the standard reference point for all
claims and reports on the subject. A further article by Francesca De
Ch^atel (2014) is sometimes also cited in support of Syria-climate
conflict arguments, but is overall quite critical of them and provides
little backing for the claims made by Femia and Werrell,
Gleick or Kelley et al. (hereafter FGK), let alone for subsequent
political statements and media reportage on the subject.
All three studies, and most other discussions of the subject, in
essence advance a three-step argument about the role of anthropogenic
climate change in the Syrian civil war: firstly, that
anthropogenic emissions were, or may have been, a contributory to
the severe late 2000s drought experienced in Syria; second, that
this drought led to large-scale internal migration; and third, that
these internal migrants were an important contributory factor in
the 2011 unrest which spiralled into Syria's civil war. It is sometimes
also claimed that increases in the price of wheat resulting
from climate change-related droughts in China and Russia were a
contributory factor to the Arab spring, in Syria included (e.g.
Sternberg, 2013) e but this ‘externalist thesis’ has received much
less attention from commentators and policymakers, and is not
discussed further here.
Before considering the various specific arguments made within
these works, we need first to be clear what is meant when it is
claimed that climate change and drought ‘helped fuel’ or ‘sparked’
or ‘contributed to’ or ‘is implicated in’ Syria's civil war. It needs
emphasising, to begin with, that despite the odd extreme statement
e such as the assertion that ‘the Syria war began because of climate
change’ by the head of a leading political risk consultancy (Channel
4, 2015) e no one seriously believes that climate change and
drought were the sole causes of Syria's civil war. Therefore simply
pointing to the existence of other causes e of which there were of
course many e does not invalidate the Syria-climate conflict thesis.
Even allowing for this, however, four distinct causal claims could be
and have been made about the role of climatic factors in the onset
of Syria's civil war: (1) that they were the final cause or spark of
conflict, as for instance in Friedman's rhetorical suggestion that ‘if a
drought is bad enough it can help push an already stressed society
to the breaking point. Is that what happened in Syria?’ (Showtime,
2014); (2) that they were the primary causal factor, as for instance in
Al Gore's claim that climate change provides ‘the underlying story
of what caused the gates of hell to open in Syria’ (Dinshaw, 2015a);
(3) that they were a significant contributory factor; and (4) that they
were a contributory factor of unknown or unspecified significance.
All four types of causal claim can be found in the literature, even
sometimes with a single report or study. For instance, Kelley et al.
(2015: pp. 3241e2) simultaneously suggest that Syria's drought
‘had a catalytic effect’ in sparking unrest; that this drought may
have been a ‘primary factor’ behind the civil war; alternatively, that
it may have been a ‘substantial factor’, or may have ‘contributed to
the conflict’; and that climate change is ‘implicated in’ Syria's civil
war. These claims are all subtly different.
For the purposes of our analysis here, such slippages also make it
difficult to know by what standard the Syria-climate conflict thesis
should be evaluated. Any critique of claim types 1, 2 and 3 would be
vulnerable to the counter that claim type 4, and therefore the entire
J. Selby et al. / Political Geography 60 (2017) 232e244 233
Syria-climate conflict thesis, continues to hold. Yet conversely, it is
impossible to demonstrate that climate change was not a factor at
all in the onset of Syria's civil war, since claims of this type (claim
type 4) are essentially unfalsifiable. Given this, instead of asking
whether climate change was actually a causal factor, of whatever
significance, in the Syrian civil war, our approach is to examine the
quality of the evidence that has thus far been marshalled in support
of this causal thesis. More specifically we ask: is there clear and
reliable evidence that climate change-related drought in Syria was a
contributory factor in the onset of the country's civil war?; and, if and
where yes, was it as significant a contributory factor as is claimed in
the existing academic and expert literature? We ask these questions
in relation to both the overall Syria-climate conflict thesis, and each
of its sub-theses. These questions place the burden of proof on the
existing literature to demonstrate the existence of clear linkages
between climate change, internal drought and Syria's civil war e
rather than on us to undertake the impossible task of demonstrating
their non-existence e whilst also leaving open the possibility
that stronger evidence may at some point emerge.
In practice, this leads us to ask several types of question of the
existing literature: about the accuracy of its data claims (e.g. was
Syria's drought the most severe on record?); about geographical
location and scale (e.g. was there a spatial correspondence between
most drought-affected areas and areas of highest out-migration?);
about temporality (e.g. to what extent did migration levels during
the drought depart from long-term trends?); about human agency
(e.g. did the conscious political demands of Syria's early demonstrators
relate to drought?); and about other intersecting causes
(e.g. were there other causes of migration in Syria besides
drought?).
In exploring these questions, our methods are as follows. We
focus primarily on the three key works introduced above, discussing
other sources only where they introduce significant additional
evidence. On the drought and its alleged links to climate change, we
analyse both Kelley et al.'s data and statistical model and raw station
rainfall data for Syria and its environs. On the alleged links
between the drought and Syria's civil war, we both revisit FGK's
sources, and draw upon evidence not considered in these studies.
We also introduce evidence from interviews with Syrian refugees
which were undertaken by one of the authors (Fr€ohlich) during
2014e15. Further details on our methods can be found in the
technical appendix.
3. Climate change and Syria's drought
FGK claim that prior to the civil war, Syria experienced a severe
multi-year drought which was partly attributable to anthropogenic
climate change. Both Femia and Werrell (2012: p. 1) and Gleick
(2014: p. 332) refer to it as ‘the worst long-term drought … since
agricultural civilizations began in the Fertile Crescent’, while Kelley
et al. call it ‘the most severe drought in the instrumental record’
(2015: 3241). All three studies refer to work by Martin Hoerling
et al. (2012) to support their contentions that Syria's pre-civil war
drought can be partly attributed to regional drying and thus to
greenhouse gas emissions. In addition, Kelley et al. perform their
own statistical modelling analysis, and conclude from this that
there has been a long-term drying trend in the Eastern Mediterranean,
that this trend is consistent with climate models, and that ‘a
drought of the severity and duration of the recent Syrian drought’
has therefore ‘become more than twice as likely as a consequence
of human interference in the climate system’ (2015: 3241).
Before interrogating these claims, two points of clarification are
in order. It deserves stressing, firstly, that none of the three key
studies above provide any data on rainfall patterns in Syria specifically.
Femia and Werrell (2012) and Gleick (2014) provide no
rainfall data whatsoever. Kelley et al.'s analysis, moreover, is of a
huge region labelled ‘the Fertile Crescent’ which includes, in
addition to Syria, territory from 14 other countries (Kelley et al.,
2015: Fig. 2). Though Kelley et al. repeatedly refer to ‘the Syrian
drought’, none of their data or statistical findings are about Syria
specifically; they are instead about this far larger Fertile Crescent
region. Hoerling et al.'s findings are likewise not about Syria specifically,
but the entire Mediterranean Basin: these authors state
explicitly that ‘the observational data are not adequate to assess
centennial changes at local scales’ (2012: 2160). Moreover, and
secondly, the three studies above offer contradictory accounts of
the duration of the rainfall shock in Syria. Femia and Werrell (2012)
refer to it as lasting from 2006 to 11, as also does Gleick (2014: p.
332); while Werrell et al. (2015: p. 32) claim it lasted from 2007 to
12, and Kelley et al. describe it as covering, variously, 2007e10
(2015: 3243), 2005e08 (2015: 3245) and 2005e10 (2015:
Figure S1). This confusion is reflected in policy and media statements,
where the drought has been described as lasting anywhere
up to six years (e.g. Mills, 2015). Greater clarity and specificity is
evidently required on the spatial and temporal scale of Syria's late
2000s drought.
Did, then, Syria and its neighbours experience a particularly
severe drought (i.e. negative rainfall shock) during the late 2000s,
and if so during which years? According to the data presented by
Kelley et al. (2015), which we have reanalysed, across the Fertile
Crescent region as a whole 2007/08 was indeed the driest year in
the instrumental record, with precipitation 35% below the 1961e90
average. Moreover, according to this same data 2006/07e2008/09
was the Fertile Crescent's driest three-year period on record, 22%
below the long-term average. Station data for northeast Syria
specifically indicates that a similar pattern applied there: in both
Qamishli and Deir ez-Zor 2007/08 was the driest year on record
(Qamishli received 25% of its 1961e90 average rainfall, Deir ez-Zor
only 12%), while 2006/07e2008/09 was the driest three-year
period on record (Qamishli received 46% of the three-year
average rainfall, Deir ez-Zor 36%). The severity of this three-year
drought was mainly a product of the extremely low rainfall in
2007/08, however: in both Qamishli and Deir ez-Zor 2006/07 and
2008/09 were dry years but no drier than many others experienced
over the last 60 years. Moreover, in 2009/10 rainfall at both stations
increased to close to average for Deir ez-Zor and above average for
Qamishli (see Figs. 2e3, and Fig. 1 for locations). This mirrors the
regional precipitation pattern: 2009/10 was an average-to-wet
winter across the Fertile Crescent, and this was followed by a
very dry winter (2010/11) and then a wet one (2011/12). These
precipitation conclusions are corroborated by analysis of the Global
Precipitation Climatology Centre's dataset conducted by Becker
et al. (2013).
The rainfall pattern across Syria during this period was far from
uniform, however. While northeast Syria witnessed extremely low
rainfall during 2007/08 and during the three-year period 2006/
07e2008/09, the same was not true in most of the rest of the
country (see Fig. 4). During both of these periods, most of western
Syria including its three largest cities, Aleppo, Damascus and Homs,
all received above average rainfall (well above average in the case of
Aleppo). And the town of Dara'a, near the Jordanian border e to
which we return below, and which was supposedly ‘especially hard
hit’, or even ‘crippled’, by the drought (Femia & Werrell, 2012;
Gleick, 2014: p. 336) e received close to average rainfall. Moreover,
it was not only Syria which experienced low rainfall during
these years. Drought also affected southeast Turkey, albeit not as
severely as northeast Syria (see e.g. data for Diyarbakir and Siirt in
Figs. 2 and 3). And it affected much of northern Iraq. Indeed, during
the two year period 2007/08e2008/09 the largest precipitation
declines were found in northern Iraq, not Syria (Trigo et al., 2010).
J. Selby et al. / Political Geography 60 (2017) 232e244234
We can now turn to the drought's causes. So, firstly, was the
drought in northeast Syria part of a long-term drying trend? Kelley
at al claim that it was, identifying a significant regional drying trend
over the period 1931e2009 (2015: 3244). Yet only 20% of the station
series that they examine across the Fertile Crescent display
significant negative rainfall trends. Moreover, Kelley et al. do
nothing to establish why their Fertile Crescent rainfall time series is
best modelled statistically as a linear trend, and nor do they present
the sensitivity of their linear trend analysis to the choice of time
period. Dryland rainfall is well known to exhibit very high natural
inter-annual and inter- and multi-decadal variability (Hulme, 1996;
Hulme et al, 2001), attributes not well modelled by linear trends.
Using linear trends to model dryland rainfall in other regions such
as the Sahel has been shown to be misleading (Dong & Sutton,
2015; Hulme, 2001). Indeed, the anomalous rainfall in the Kelley
et al. (2015) Fertile Crescent region is very largely focused on the
decade 1999e2008, which was 10% drier than the long-term
average; while Hoerling's et al. (2012) analysis of the wider Mediterranean
picks out the decade 1985 to 1995 as the driest. In both
studies, the linear trend for the majority of the twentieth century is
weak to non-existent. Rather than characterising rainfall in the
Fertile Crescent as displaying a long-term drying trend, the
observed rainfall series is better characterised as highly variable
with an anomalous decade from 1999 to 2008, culminating in a
severe drought (centred on 2007/08) after which rainfall returned
to its pattern of high inter-annual variability (wet, dry, wet). Our
analysis of raw station data from northeast Syria confirms this
finding (see technical appendix). We conclude that there is no
evidence of progressive multi-decadal drying either in the Fertile
Crescent region as a whole, or in northeast Syria specifically.
Last, can the drought be attributed to human influences on the
global climate system? The science of extreme weather event
attribution is still young and subject to competing methods and the
constraints of limited observational data and imperfect models
(Bellprat & Doblas-Reyes, 2016; Hulme, 2014). The attribution
method used by Kelley et al. effectively involves three steps:
identifying a long-term drying trend, as discussed above; estimating
the increased likelihood of the drought occurring given this
trend; and comparing the trend to results from climate model
simulations. There are two problems here. On the one hand, given
that there is no evidence of a long-term drying trend either in the
Fertile Crescent region as a whole or, more importantly, in Syria
specifically, as argued above, it follows that claims about the
increased likelihood of severe droughts occurring because of a
claimed trend are invalid. And on the other hand, the rainfall
simulations on which Kelley et al. (and also Hoerling et al.) rely
involve huge uncertainties (see technical appendix). It is also worth
noting that, even if Kelley et al. were right that human influences
had caused the frequency and severity of droughts in Syria to increase,
it would still not follow that these human influences were
causally ‘implicated’ in this particular drought (2006/07e2008/09):
to claim thus, as Kelley et al. (2015: 3241) do, is to confuse probabilistic
and deterministic claims (Solow, 2015).
In summary, northeast Syria did experience a severe drought
during the three years 2006/07e2008/09, and unprecedentedly so
during 2007/08. However, contrary to what Femia and Werrell,
Homs
Siirt
Dara'a
Aleppo
Damascus
Qamishli
Diyarbakir
Deir ez-Zor
0 110 22055 km
TURKEY
SYRIA
IRAQ
JORDAN
LEBANON
ISRAEL
Hasakah
Governorate
Fig. 1. Syria and its neighbours. The figure identifies all locations (cities, rainfall stations, governorates, and neighbouring states) discussed in the text.
J. Selby et al. / Political Geography 60 (2017) 232e244 235
Fig. 2. Percentage anomalies of ‘annual’ November-April rains, relative to 1961e90 average, for Qamishli (Syrian/Turkey border), Deir ez-Zor (northeast Syria) and Diyarbakir
(southern Turkey). Years are dated by the November of this winter rains period. The last year shown is for 2010/11. Data source: David Lister, Climatic Research Unit, University of
East Anglia.
Fig. 3. Percentage anomalies of ‘annual’ November-April rains, relative to 1961e90 average, for the period 2000/01 to 2010/11 for five stations in Syria and Turkey. Data source:
David Lister, Climatic Research Unit, University of East Anglia.
J. Selby et al. / Political Geography 60 (2017) 232e244236
Gleick and others claim or imply, this was not a five-plus year
drought, and did not affect the whole of Syria. Pace Kelley et al., our
analysis also questions whether the historic pattern of rainfall in
the Eastern Mediterranean is best described as a ‘long-term drying
trend’. In view of both this and the uncertainties in climate model
simulations, we therefore conclude that there is no clear and reliable
evidence that human influences on the climate system are
implicated in northeast Syria's 2006/07e2008/09 drought.
4. Drought and migration
Step two of the Syria-climate conflict thesis involves the claim
that Syria's drought led to ‘total crop failure’ (Femia & Werrell,
2012), ‘agricultural failures’ (Gleick, 2014: p. 334), and maybe
even wholesale agricultural ‘collapse’ in the northeast governorate
of Hasakah (Kelley et al., 2015: p. 3241), and that this in turn
precipitated large-scale migration from the Syrian countryside, and
from Hasakah in particular. Kelley et al., Gleick and Femia and
Werrell all make this case, in only slightly different ways: thus
Kelley et al. claim that ‘as many as 1.5 million people’ were ‘internally
displaced by the drought’ (2015: 3241e2), Gleick that ‘more
than 1.5 million people’ migrated (2014: 334), and Femia and
Werrell (2012) that there was ‘a massive exodus … to the cities’.
In subsequent media reports, this 1.5 million figure has often
morphed into the claim that up to 2 million Syrians were displaced
by the drought (including by Femia and Werrell: see e.g. Baker,
2015; Werrell & Femia, 2015).
In evaluating this sub-thesis, it is worth acknowledging from the
outset that there is abundant evidence that Syria's 2006/07e2008/
09 drought did have some significant agricultural, rural livelihood
and migration effects, especially within the governorate of Hasakah.
UN drought impact assessments, Syrian government statements,
US diplomatic cables and international media reports all
corroborate this, as do the interviews used in the present study.
There would not have been successive international drought appeals
(UN-OCHA, 2009a, 2008) in the absence of significant
drought impacts. Indeed, we know of no evidence and see no
reason to think that exceptionally poor rains were not a contributory
factor in internal migration during 2008e09. Hence the relevant
analytical question becomes not whether there does exist
evidence of a link between drought and migration e since there
does e but whether this link was of the scale and significance that
is claimed by FGK and others.
Here the obvious starter question is whether the claimed
migration numbers are at all plausible. It is of course impossible to
say precisely how many people migrated from northeast Syria
during 2008e09, especially as no comprehensive survey of migrant
numbers was ever undertaken. What can be said with confidence,
however, is that there is no meaningful foundation for a 1.5
let alone 2 million figure. The sole source for the former was a short
humanitarian news report, citing a single Syrian government officer
as claiming that 1.25e1.5 million people had been displaced by
the drought (IRIN, 2009b). Our best guess is that this figure was
meant to refer to the number of people affected by e not displaced
by e the drought, given the UN's estimate for numbers affected of
1.3 million (UN-OCHA, 2009a: 1). In any case, the 1.5 million figure
claimed by Gleick and Kelley et al. is completely out of line with
Syrian government and UN estimates, and all other sources of
which we are aware. The first UN drought appeal, launched in
September 2008 following the extreme 2007/08 drought, gave no
figure at all, claiming only that migration from rural areas during
2007/08 was 20e30% higher than during previous years (UNOCHA,
2008: p. 1). By summer 2009 the UN, using Syrian government
figures, was estimating that 40e60,000 families had migrated
because of the drought, including 36,000 families, or 200e300,000
people, from Hasakah (UN-OCHA, 2009a: p. 5; UN-OCHA, 2009b:
Fig. 4. Percentage anomalies of ‘annual’ rainfall for the three-year period 2006/07-2008/09 relative to the 1982/83-2009/10 average for Syrian stations. Blue is wetter than average;
orange/red is drier. Data source: MAAR, 2016. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
J. Selby et al. / Political Geography 60 (2017) 232e244 237
pp. 18e19). A subsequent UN report put total drought migration at
65,000 families (UN-OCHA, 2010: 6), plus the UN Special Rapporteur
at one point claimed 600,000 people had migrated (De
Schutter, 2010); but 40e60,000 families remained the most
widely used figure (e.g. IRIN, 2009a; Massoud, 2010). Even the
highest of these figures is far below the 1.5 or 2 million suggested
by FGK.
Numbers aside, the other key question to ask is: was the 2006/
07e2008/09 drought as significant a cause of migration from
northeast Syria as is claimed within the existing literature? To be
sure, FGK recognise that the pre-civil war out-migration from
northeast Syria was not caused by poor rains alone, and to this
extent none of them offers a mono-causal account of the link between
drought and migration. Instead, they each emphasise that
‘agricultural collapse’ and the ensuing migration were caused by a
‘complex interplay of variables’, which included declining
groundwater levels, inefficient irrigation, over-grazing, population
growth, and poor governance (Femia & Werrell, 2012; 2015: p. 33;
Gleick, 2014: pp. 334e6; Kelley et al., 2015: pp. 3241, 3243). These
emphases are in keeping with research showing that vulnerability
to environmental shocks is caused by a multiplicity of environmental,
economic and governance factors (e.g. Adger, 2006), and
that ‘environmental migration’ is never caused by environmental
variables alone (e.g. Black et al., 2011). Nonetheless, it still needs
asking whether FGK provide accurate accounts of the links between
drought and other variables in the making of northeast Syria's precivil
war migration crisis. In our assessment, there are important
respects in which they do not.
To start with, several crucial variables are neglected by FGK.
Their accounts include barely any mention, most strikingly, of the
dramatic economic transformations that Syria had been undergoing
prior to and during the 2006/07e2008/09 drought (two short
sentences in Kelley et al., [2015: p. 3242] excepted: we return to
these below). For, starting in the 1990s and intensifying following
Bashar al-Assad's accession in 2000, Syria had made a decisive turn
towards the market economy, including through the privatisation
of state farms, trade liberalisation, and the removal of key subsidies
e including, in May 2008, the removal of fuel subsidies (which led
to an overnight 342% spike in fuel prices) and, in May 2009, the
removal of fertilizer subsidies (which led to price increases of
200e450%: De Schutter, 2011: p. 16).
FGK also fail to note that large-scale migration from and within
rural Syria, and especially from Hasakah, was occurring well before
the 2006/07e2008/09 drought. There had long been a pattern of
seasonal agricultural labour migration, involving men from northern
and northeast Syria spending the summer working in the
south, before returning to their families and landholdings in the
autumn. Livestock herding had also long involved extensive
migration, with herds typically spending their winters on the open
steppe, before being transported as far as the Mediterranean in
search of crop residues or greener pastures (Leybourne, 1998: pp.
201e2). More importantly, the liberalisation programme discussed
above led to a sharp rise in rural-to-urban migration: between
2000 and 2005, an estimated 135,000 people were leaving rural for
urban areas each year (estimated from al-Hindi, 2011: p. 25), and
even prior to the drought the peri-urban areas and informal
housing in and around Syria's cities were growing rapidly
(Goulden, 2011). Hasakah specifically was already witnessing net
annual out-migration estimated at 1.7% during the late 1990s
(Khawaja, 2002: p. 27), while farmland in Hasakah was already
being extensively abandoned before the 2006/07e2008/09 drought
(Hole, 2009).
This leads us to the key issue, which is that FGK's inattention to
Syria's changing political economy leads them to systematically
overstate the impacts, both direct and indirect, of the 2006/
07e2008/09 drought on migration. With respect to the indirect
effects, firstly: FGK repeatedly misattribute important socioeconomic
changes within Syria e changes which, they imply,
contributed to the migration from the northeast e to the drought,
when in fact they were long-term corollaries of the country's
troubled liberalisation. For example, while Kelley et al. (2015: p.
3241) imply that it was drought which led to a sharp decline in
agriculture's contribution to Syrian GDP between 2003 and 2008,
this decline in fact predated the 2006/07e2008/09 drought, and
non-drought years, such as 2003 and 2004, experienced amongst
the steepest declines (World Bank, 2016). Kelley et al. (2015: p.
3242) also claim that livestock feed and food price inflation in Syria
prior to the civil war was mainly caused by drought; but this cannot
have been more than a secondary factor given that the entire Syrian
economy was plagued by high inflation during this period (during
2007/08, for instance, the cost of utilities rose by an estimated 87%:
US Embassy Damascus, 2008c). Femia and Werrell (2012) and
Gleick (2014: p. 334) claim that the drought drove two to three
million people into food insecurity and extreme poverty; and yet
the UNDP estimates on which these claims are based (El Laithy &
Abu-Ismael, 2005; El Laithy & Abu-Ismael, 2009) were for predrought
extreme poverty levels (2 million during 2003e04; 2.4
million during 2006e07). In the one place where these various
authors make any mention of Syria's experiment with liberalisation,
it is to imply that Bashar al-Assad ‘shifted to liberalizing the
economy’ in response to the drought (Kelley et al., 2015: p. 3242) e
a suggestion completely at odds with the historical record. Indeed,
contrary to what FGK imply, the steep decline in agricultural
employment in Syria during the pre-civil war years was ‘clearly
linked to major structural transformations of the sector, more than
to drought and climatic conditions’ (Aita, 2009b: p. 76; see also;
Ababsa, 2011). While the drought would no doubt have had socioeconomic
consequences, it is clear that many e perhaps even most
e of FGKs claims about these consequences are overstated and
unreliable.
The same applies to FGK's portrayals of the final causes of the
2008-09 migration from northeast Syria. FGK assume that, whatever
the contextual factors that may have increased Syria's
vulnerability to shocks, it was an exceptionally severe drought
which ultimately compelled people to migrate. And yet drought
was not the only shock event to affect northeast Syria during
2007e09. As already noted, rural Syria was hit during both 2008
and 2009 by overnight triple-digit increases in the price of crucial
agricultural inputs, following the removal of government subsidies.
In addition, 2007 saw the implementation of a new agrarian relations
law giving landowners the right to terminate tenant contracts
with immediate effect, while 2008 saw the enactment of
Presidential Decree No. 49, which restricted land sales across the
whole of Hasakah governorate (Ababsa, 2015: pp. 210e11, 216e17).
During 2008, Syria's wheat crop was also hit by exceptionally severe
late-winter frost. None of these other ‘shock factors’ can be
dismissed as irrelevant. The view of the US Embassy in Damascus
was that Hasakah's poor wheat harvest during 2008 was ‘mostly
due’ to the late frost, with drought having been merely one of
several ‘contributing factors’ (US Embassy Damascus, 2008a). De
Ch^atel, in her analysis of the Syria climate-conflict thesis, claims
that ‘for many farmers in the Jazira and elsewhere’ the cancellation
of subsidies ‘formed a greater burden than the successive years of
drought and spurred their decision to abandon their land’ (2014:
526). And Decree No. 49 has been cited as a key factor in the
increased Kurdish out-migration from northeast Syria pre-civil war
(ACCORD and DIS, 2010). At the very least, it should not be assumed
that drought was the only, or dominant, catalyst of migration.
Consideration of the economic structure of agriculture in
northeast Syria suggests an even stronger conclusion, however. For,
J. Selby et al. / Political Geography 60 (2017) 232e244238
except in the case of purely rain-fed crops such as barley, most
agricultural production in the northeast, including in Hasakah, was
not that dependent on good rains. The wheat and livestock sectors
both illustrate this. For obvious reasons, wheat yields are strongly
affected by drought when rainfed, but relatively stable when irrigated;
and given that over 60% of Syria's wheat production and the
vast majority of Hasakah's production was irrigated, it follows that
yields should not have been highly susceptible to drought
(Westlake, 2003: pp. 140e41). By contrast, as was observed prior to
the price liberalisation, ‘most wheat farmers … would be unable to
sustain production if the government were to withdraw its price
support’ (Westlake, 2003: p. 162). And similarly for livestock:
because Syrian sheep flocks are primarily fed on artificial supplements
and crop residues, including from irrigated areas, and obtain
as little as 9% of their food requirements from natural fallow-land
and range grazing, ‘[t]he stability of the system is far more
dependent upon the availability of supplements and the price of
feedstuffs than upon the availability of natural forage’ (Jaubert &
Bocco, 1998: p. 188). Assuming this is correct, then the sudden
removal of fuel and fertilizer subsidies in 2008 and 2009 must have
had greater immediate impacts on farmer livelihoods and migration
than the drought. The fact that government- and UNsponsored
‘drought relief’ efforts were concentrated, paradoxically,
in an area of Hasakah where rainfall levels are so naturally low
that wheat production is dependent on groundwater irrigation, and
not in areas of rain-fed agriculture (UN-OCHA, 2010: p. 4; De
Schutter, 2011: p. 12), lends further support to this conclusion.
Overall we conclude that, while there is no reason to doubt that
the 2006/07e2008/09 drought did increase migration from
northeast Syria during 2008e09, FGK significantly overstate both
the scale of the migration, and the extent to which it was caused by
drought. The official UN and Syrian government estimate was that
40e60,000 families migrated from northeast Syria during
2008e09, not the 1.5 or 2 million people that is often claimed.
Moreover, given that large-scale migration from northeast Syria
predated the drought, it follows that ‘excess migration’ (i.e.
migration during 2008e09 minus average pre-drought migration
numbers) must have been lower still. Numbers aside, we also find
that FGK present a mistakenly environment- and drought-centric
reading of the causes of northeast Syria's agrarian and migration
crisis, and that other factors, most importantly Syria's experiment
with economic liberalisation, were likely more important contexts
for and catalysts of migration than the drought.
5. Migration and civil war
We can now turn to the final step of the Syria-climate conflict
thesis, namely that migrants from the drought in northeast Syria
were a contributory factor in the country's 2011 unrest. FGK claim
essentially this: that the sudden presence of large numbers of rural
poor ‘placed significant strains on Syria's economically-depressed
cities' and provincial towns, and became one ‘driver’ of unrest
(Femia & Werrell, 2012); that drought-related migration ‘contributed
to urban unemployment and economic dislocations and social
unrest’ and ‘spurred many Syrians to make their political grievances
publicly known’ (Gleick, 2014, pp. 333, 336, quoting; Saleeby,
2012); and that ‘population shock’ to Syria's urban peripheries
‘exacerbated a number of the factors often cited as contributing to
the unrest’, including ‘unemployment, corruption, and rampant
inequality’ (Kelley et al., 2015: p. 3242).
In considering the merits of this thesis, it is worth emphasising
to start with just how meagre the evidence for it is. Femia and
Werrell (2012) devote only one paragraph to the issue, Kelley
et al. (2015: p. 3242) just two, and Gleick (2014: pp. 333e5), in
the most extensive treatment of the links, just a couple of pages. No
attempt is made within these studies either to statistically correlate
climatic and conflict variables or to make use of original interview
or ethnographic research; instead all three studies rely exclusively
on journalistic and policy advocacy sources. Moreover, none of
them refers to the body of academic literature on the causes of
Syria's civil war. Particularly striking in this regard is Kelley et al.'s
analysis, which devotes considerable effort to establishing a
connection between climate change and the Fertile Crescent
drought, while relying on secondary and journalistic sources to
establish a link between drought and Syria's conflict. Indeed, the
little evidence that is provided by FGK and others in support of the
drought migration-civil war thesis is essentially threefold,
comprising: claims that drought-induced migration was a key
element in the ‘population pressures’ within pre-civil war Syria;
evidence on the timeline and geography of Syria's early unrest; and
testimony from individual Syrians. We examine each of these types
of evidence in turn.
Firstly, then, was drought-related migration a key element in the
‘population pressures’ within pre-civil war Syria (and should it be
assumed that these pressures fed unrest)? Kelley et al. contend just
so, implying that Syrians displaced by drought accounted for
around half of Syria's 2003e10 urban growth (2014: Figure S1), and
that the ‘conflict literature supports the idea that rapid demographic
change encourages instability’ (Kelley et al., 2015: p.
3242). Yet neither contention is correct. The major (positive as well
as negative) population changes in Syria during this period were:
(1) natural population growth of up to 3% annually: around 3
million (Aita, 2009a: pp. 10e11); (2) the arrival of Iraqi refugees: 1.5
million (Kelley et al., 2015); (3) general rural-to-urban migration:
around 1 million (extrapolated from al-Hindi, 2011: p. 25); (4) outmigration
from Syria, estimated at 128,000 annually: 900,000 (Aita,
2009a: p. 47); (5) the end of circular migration to Lebanon in 2005:
several hundred thousand (Chalcraft, 2009); and (6) ‘excess
migration’ from Syria's northeast during 2008e09: a proportion of
the estimated 40e60,000 families. Given, in addition, that this
‘excess migration’ is not wholly attributable to drought, as discussed
above, it is evident that Kelley et al. significantly overstate
the contribution of drought-related migration to Syria's pre-civil
war urban growth (see technical appendix for further calculations).
In any case, it should not be assumed that these population
changes contributed to unrest. Pace Kelley et al., while some studies
do indeed find a link between demographic change and conflict
(Goldstone, 2002), others have found to the contrary (e.g. Urdal,
2005; Buhaug and Urdal, 2013). Moreover, while some studies
have found links between environmental migration and conflict
(e.g. Reuveny, 2007), more have disputed this, including by finding
that migration is a mechanism of adaptation and thus conflict
mitigation (e.g. Black et al., 2011; Hartmann, 2010). The fact that
there were significant demographic and migration changes in Syria
prior to the civil war in itself lends no weight to the Syria climate
conflict thesis. If such changes did play a causal role, then this needs
to be demonstrated empirically.
Second, then, does the chronology and geography of Syria's early
unrest support the drought migration-civil war thesis? FGK aver
that it does, claiming that Syria's ‘rapidly growing urban peripheries’
(Kelley et al., 2015: p. 3242) as well as ‘disaffected rural
communities’ (Femia & Werrell, 2012) were at the heart of the
developing unrest; and that the provincial town of Dara'a, where by
most accounts the civil war began, had been ‘especially hard hit by
five years of drought’ (Femia & Werrell, 2012) and had witnessed ‘a
particularly large influx of farmers and young unemployed men
displaced off their lands by crop failures’ (Gleick, 2014: p. 335). Are
these claims convincing?
It is indeed generally accepted within the literature on the
Syria's 2011 uprising that its geography was much more ‘rural and
J. Selby et al. / Political Geography 60 (2017) 232e244 239
rurban’ (De Ch^atel, 2014: p. 521) than in the parallel case of Egypt
(see esp. Barout, 2011); and it is also widely accepted that Dara'a
was ‘where the Syrian uprising began’ (Leenders & Heydemann,
2012: p. 140). In fact, the first post-Arab Spring protests in Syria
took place, not in Dara'a, but on 17 February in the old commercial
district of Al-Hariqa in central Damascus, following a police officer's
assault on the son of a shop owner. The second protests also took
place in Damascus, on 22 and 23 February, when a group of youths
held a vigil outside the Libyan embassy in solidarity with opponents
of Muammar Qadhafi. Protests were also ongoing in Kurdish
areas of the northeast. Mass protests did not begin elsewhere,
including in Dara'a and then other provincial towns, until midMarch
2011. None of this in question.
What is in question, however, is whether the undoubted centrality
of Dara'a and other provincial areas to the uprising corroborates
the drought migration-civil war thesis. There are two issues
here. Firstly, it is not clear that the uprising took hold in rural and
provincial areas for reasons of socio-economic determination.
Though some have argued this (see esp. Barout, 2011: pp. 21e22),
statistical evidence is contradictory, with some research finding, for
instance, that Dara'a governorate experienced particularly acute
increases in poverty pre-civil war (SCPR, 2013: p. 21), and other
research characterising Dara'a as a ‘low deprivation’ region which
had experienced ‘significant improvement’ in poverty levels
(UNICEF, 2014: p. 41). Indeed, the most in-depth analyses of the
origins of Syria's civil war suggest that the uprising took hold in
rural and peripheral areas such as Dara'a, not for socio-economic
reasons, but because of patterns of regime repression and political
opportunity (Leenders, 2012; Leenders & Heydemann, 2012).
Initially during early 2011, the regime focused its attention mostly
on urban (and Kurdish) areas, where unrest was thought most
likely; in consequence, security and police forces were redeployed
away from peripheral areas, including Dara'a. The outbreak of mass
protests in this supposedly quietist town thus caught the regime by
surprise. It was within this context that, from 22 March, state security
forces unleashed a level of violence against protesters in
Dara'a which was beyond anything seen to that point, with regime
repression providing the impetus for further local mobilisation
(Leenders & Heydemann, 2012: pp. 142e3). Dara'a also benefitted
from its location adjacent to Jordan, which enabled protestors to
utilise established cross-border and migration networks for arms,
funding and escape. Syria's civil war, by such interpretations, was
not initially a ‘rural and rurban’ uprising, but instead became one as
a result of regime repression combined with the opportunities
afforded by rural and peripheral location.
Moreover, even if economic marginalisation and grievances
were a factor in the geography of the uprising, this would still not
constitute evidence of the impact of drought-related migration
specifically. There were, as discussed above, multiple sources of
rural socio-economic grievance in pre-civil war Syria, besides
drought e the long-term process of agricultural liberalisation, the
ensuing long-term pattern of rural out-migration, and the sudden
removal of input subsidies, most notably. And Dara'a had not been
‘especially hard hit’ by drought, as noted above. While this does not
preclude the possibility that the presence of migrants from the
northeast may have been a factor in anti-regime mobilisation in
Dara'a e a possibility to which we return below e existing scholarly
accounts do not mention or consider it (Barout, 2011; Leenders,
2012). In our assessment, then, the timeline and geography of
Syria's early unrest does little to corroborate the drought
migration-civil war thesis.
What, lastly, of testimony from individual Syrians? ‘The drought
and unemployment were important in pushing people towards
revolution’, says a Syrian farmer quoted by Kelley et al. (2015: p.
3245, following Friedman, 2013); ‘the drought was beyond our
capacity as a country to deal with’, claimed the Syrian Minister of
Agriculture as quoted by Gleick (2014: p. 334, from US Embassy
Damascus, 2008b). Clearly, such comments should not be dismissed
as irrelevant, since subjective representations and
reasoning are critical elements in the descent into war, as in any
political process. That said, it is striking how little personal testimony
is marshalled in support of the claim that drought-related
migration was a factor in the Syrian uprising: Kelley et al. quote
just one displaced Syrian farmer, Gleick and Femia and Werrell
none at all.
In sum, the evidence marshalled by FGK on the link between
drought migration and civil war is extremely weak. Even more
striking than this, however, is what their accounts omit: namely,
any consideration of whether migrants from northeast Syria were
significantly involved, whether as mobilisers, participants or targets,
in the early demonstrations which spiralled into civil war.
Neither FGK, nor to our knowledge anyone else, has provided any
evidence of this. And there are at least two reasons that suggest to
the contrary.
First, the interviews with Syrian refugees conducted by one of
this study's authors suggest that, in Dara'a at least, migrants from
the northeast did not participate to any significant degree in the
spring 2011 protests, were not targets of the demonstrations or
subsequent repression, and left as soon as the protests started and
went back to the northeast. ‘They had nothing to do with politics.
They went to work and back home’ (interview Zaatari, 6 January
2015, interviewee from southern Dara'a). ‘They had no opinion.
Their life revolved solely around their work’, said one (interview
Zaatari, 6 October 2014a; interviewee from Dara'a). ‘In the first year
some of them stayed. But when the troubles grew, they left’
(interview Ramtha, 10 October 2014; interviewee from Tafas,
Dara'a). ‘They left at the beginning of the demonstrations’ (interview
Zaatari, 7 January 2015, interviewee from Dara'a). Indeed,
most of the migrants from the northeast seem not to have even
lived in Dara'a city, where the uprising began, but rather in tents
and small houses scattered throughout surrounding rural areas
(Ababsa, 2015: p. 210; Fr€ohlich, 2016). Of the 32 interviews conducted
on the subject, in only one did the interviewee claim that
migrants from the northeast were at all involved in Syria's early
unrest.
Second, none of the political demands made by Syria's early
2011 protest movements related directly to either drought or
migration. If migrants from the northeast had been significantly
involved in the early unrest, whether as subjects or objects, then
this would have been reflected within political discourse and demands
(e.g. in demands for them to be sent home, or for host
communities to receive increased government support). Yet the
February-March 2011 protests in Damascus focused on denouncing
the security forces and the broader apparatus of authoritarianism,
including demands for the release of political prisoners; and
reflecting this, the two main slogans of the protest movement
became ‘al-sha'ab as-souri ma biyinthal’ (‘the Syrian people will not
be humiliated’), and ‘yalli bi-eqtul sha'bo khayen’ (‘he who kills his
own people is a traitor’). In Dara'a, likewise, initial demands
revolved around civil rights and political freedoms, plus economic
liberalisation-related grievances. For example, the 13 demands
presented in March 2011 by Dara'a notables and representatives of
aggrieved families to the President's emissary, Major General
Hesham Ikheteyar, included: removal from their offices of the head
of the local Political Security branch and the town Governor;
apologies to the families who had been offended by the Governor;
political reform and greater political freedoms; a meeting with the
President; the return of women teachers who had been expelled
following introduction of a law banning them from wearing the
niqab; the expulsion from the town of all companies owned by
J. Selby et al. / Political Geography 60 (2017) 232e244240
Rami Makhlouf (the President's cousin and country's leading
business magnate); the removal of new laws regarding land sales;
and the lowering of the price of fuel (Barout, 2011: pp. 186e7). The
last three of these demands related to grievances arising from
Syria's troubled economic liberalisation. By contrast, we know of no
slogans, demands or chants made in Dara'a which can be directly or
indirectly attributed to the presence of migrants.
In summary, the evidence marshalled by FGK and others in
support of drought migration-civil war thesis is extremely weak:
neither their assertions about population pressures, nor their
claims about the chronology and geography of Syria's early unrest,
nor indeed the testimonies they quote, provide any firm basis for
concluding that migration from northeast Syria was a factor in civil
war onset. To the contrary, our evidence suggests that migrants
from the northeast were not significantly involved in the early 2011
unrest.
6. Conclusions
In light of the above we can now return to our main questions: is
there clear and reliable evidence that climate change-related drought
in Syria was a contributory factor in the onset of the country's civil
war?; and, if and where yes, was it as significant a contributory factor
as is claimed in the existing academic and expert literature? On each
step of the claimed causal chain, our answers are no. We find that
there is no clear and reliable evidence that anthropogenic climate
change was a factor in northeast Syria's 2006/07e2008/09 drought;
we find that, while the 2006/07e2008/09 drought in northeast
Syria will have contributed to migration, this migration was not on
the scale claimed in the existing literature, and was, in all probability,
more caused by economic liberalisation than drought; and
we find that there is no clear and reliable evidence that droughtrelated
migration was a contributory factor in civil war onset. In
our assessment, there is thus no good evidence to conclude that
global climate change-related drought in Syria was a contributory
causal factor in the country's civil war.
This finding, to be clear, does not prove that climate change and
northeast Syria's drought were not factors in its civil war, only that
the existing claims to this effect do not stand up to close scrutiny. It
remains possible that additional and stronger evidence of a humaninduced
long-term drying trend in northeast Syria, or of linkages
between drought migration and Syria's unrest, may emerge; plus it
deserves stressing that nothing in the above refutes the ‘externalist
thesis’ about the links between climate change and Syria's civil war.
Nonetheless, as in the case of Darfur, these findings suggest the
need for far greater caution and rigour, and lend further weight to
critiques of ‘threat multiplier’ climate conflict discourse (e.g.
Hartmann, 2010; Selby and Hoffmann, 2014b). For, contrary to the
claims of its advocates, ‘threat multiplier’ discourse is neither
cautious nor rigorous, instead typically combining dystopian
speculation and exaggerated accounts of the likely impacts of
climate change (see esp. CNA Military Advisory Board, 2007),
together with periodic, if contradictory, calls for the avoidance of
over-simplification (e.g. Werrell & Femia, 2015). Given the urgency
of the climate change challenge and the contestation around it, plus
the mass media's preference for striking, overblown stories e as is
clearly illustrated by the take-up of Kelley et al. (2015) e it is in our
view incumbent on analysts not to exaggerate climate-conflict
linkages, or to champion false but headline-friendly statistics.
Equally, given the complexities involved in analysing the socioeconomic
and political impacts of climate change e as clearly
demonstrated above e it is crucial that researchers working on
these impacts draw upon cross-disciplinary expertise, and not, as
Femia and Werrell, Gleick and Kelley and colleagues do, expound
on the causes of migrations and conflicts without even drawing on
social scientific research on the subjects. The case for international
action on climate change is strong enough without resort to
dubious evidence of its impact on the Syrian civil war.
Conflict of interest statement
The authors of this manuscript have no conflicts of interest.
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
Acknowledgements
Thanks to Katy Joyce and Wassim Naboulsi, School of Global
Studies, University of Sussex for their research assistance; to Mark
de Jong, Department of Geography, King's College London, for
preparing the climate figures; to the four anonymous referees; and
to Myriam Ababsa, Michael Brzoska, Ciro Fiorillo, Priya Gujadhur,
Reinoud Leenders, David Lister, Kamran Matin, Theodore Murphy,
Sara Pavanello and the Syrian Center for Policy Research for their
assistance, comments and advice. Christiane Fr€ohlich’s field
research used in this article was funded by the DFG Cluster of
Excellence “Climate System Analysis and Prediction” (CliSAP) at
Hamburg University.
Technical appendix
The geographical extent of Kelley's et al. and Hoerling's et al.
analyses
Kelley et al. (2015) analyse rainfall trends across what they label
the ‘Fertile Crescent’ region. They define this region as including, in
addition to Syria, all of Armenia, Lebanon and the Palestinian territories;
most of Azerbaijan, Cyprus, Iraq, Israel, Jordan; large parts
of Iran and Turkey; and small areas of Egypt, Georgia, Russia, Saudi
Arabia (Kelley et al., 2015: Fig. 2) Kelley et al. examine rainfall data
from 25 stations across this Fertile Crescent region, of which only a
small fraction are located in Syria. Hoerling et al. (2012) treat the
entire Mediterranean Basin as their region of analysis; their analysis
includes no stations from Syria (see their Fig. 2: p. 2149).
Rainfall data methodology
Kelley et al.'s rainfall data is measured by the important winter
rains period, November-April, which delivers the vast majority of
annual rainfall. It is for this reason that our analysis refers to rainfall
years (e.g. 2007/08) rather than calendar years (e.g. 2007). 2008/09
is the last year analysed by Kelley et al., but we also draw on their
data for 2009/10 and 2010/11 (data supplied by Colin Kelley to
authors, 4 October 2015).
For Syria specifically, we analyse two sets of data: (1) monthly
rainfall series for all Syrian stations contributing to Climatic
Research Unit (CRU) gridded data products (Harris et al., 2014).
There are 14 such data series in the CRU station archive, the longest
(Damascus) commencing in 1911, seven more commencing in the
inter-war years, and the remainder commencing only since 1945.
Only seven of these series have archived data after 1998. Just two of
these seven CRU station series are in the northeast of Syria, the area
claimed as the source of Syria's pre-civil war migration: Qamishli
on the border with Turkey (commencing 1952), and Deir ez-Zor on
the Euphrates (commencing 1946). We analyse these two station
series in the same way that Kelley et al. do for the aggregated series
for the FC region. Both of these station series have several months
of missing data (about 20% of the months in question) for the years
2004e2012. The missing data is estimated using percentage deviations
from the 1961e90 average for the remaining months in
J. Selby et al. / Political Geography 60 (2017) 232e244 241
each respective rainy season (i.e. it was assumed that missing
months' percentage total was identical to the percentage total of
preceding and following months). We also analyse (2) annual
rainfall year data from the Syrian Ministry of Agriculture and
Agrarian Reform (MAAR, 2016). This data covers 54 stations for the
period 1982/83e2009/10. Fig. 4 uses this data.
Analysis of rainfall trends
Of the 25 rainfall station series examined by Kelley et al., only
five display significant linear drying trends (one displays a wetting
trend). Moreover, these trends are of marginal significance (p < 0.1,
rather than the more conventional p < 0.05). Only one of these five
stations is in Syria (Deir ez-Zor).
For the Fertile Crescent as a whole, the linear drying trend is
weak to non-existent for the majority of the twentieth century
(1931e1998); it is far from being significant (p ¼ 0.27).
Analysis of raw station data from northeast Syria confirms these
findings. Subjecting the Qamishli and Deir ez-Zor time series to
linear regression, as performed by Kelley et al., shows that although
rainfall declined on average through their 60 (Qamishli) and 65
(Deir ez-Zor) year periods, this decline was nowhere near significant
for the former station (p ¼ 0.72) and only close to significant at
the 5% level for the latter (p ¼ 0.06). In line with our argument
above, moreover, the trends over the twentieth century (up to
1998/99) are even less significant (p ¼ 0.30 and p ¼ 0.34 respectively).
In short, there is no significant ‘long-term drying trend’ in
northeast Syria over the recorded series as captured by linear
regression. Rather, the behaviour of rainfall in this region e
consistent with what is known about semi-arid rainfall regions
elsewhere e is for high inter- and multi-annual variability, punctuated
by occasional very severe droughts and dry decades, of
which the 2000s was one.
Use of climate model simulations
The detection and attribution method used by Kelley et al. detrends
the observed time series before applying a frequency analysis
to the residual extremes and comparing this to results from
climate model simulations. Specifically, they use an ensemble of
CMIP5 models in simulating rainfall over the Fertile Crescent region.
There are several problems with using such models, not least
that individually they contain biases of up to 40% in both climatology
and inter-annual variability (Kelley et al. Fig. S3). Physical
understanding of regional rainfall responses to human influences
remains incomplete. For example, recent analysis by He and Soden
(2017) questions the reliability of the sub-tropical precipitation
decline thesis, suggesting that in response to human influences on
the atmosphere rainfall declines in the sub-tropics are more likely
over ocean areas (e.g. over the Mediterranean Sea) than over land
(e.g. northeast Syria). Moreover, the logic involved in using a model
‘ensemble of opportunity’ rather than an ensemble designed according
to model independence criteria (Sanderson et al., 2015) is
uncertain. Kelley et al. claim (2015: SI) that the CMIP5 ensemble of
opportunity projects a reduction in future winter rainfall in the FC
region. While true, the strength of ‘agreement’ within this
ensemble is weak and not significant relative to the standard deviation
of the model-estimated natural variability of 20 year means
(Stocker et al., 2014: p. 1356, Figure AI.42).
The contribution of ‘excess migration’ from northeast Syria during
2008e09 to Syria's 2003e10 urban growth
As a very rough approximation, on the basis of figures in the
main text we estimate Syria's urban population growth during
2003e10 at 3.9e4.2 million, as a result of: (1) natural urban population
growth: þ1.5 million (50% of Syria's total estimated population
growth); (2) the arrival of Iraqi refugees: þ1.5 million (100%
of the Iraqi refugees); (3) general rural-to-urban migration: þ1
million; (4) out-migration from Syria: À450,000 (50% of Syria's
total estimated out-migrants); (5) the end of circular migration to
Lebanon in 2005: þ200,000; and (6) ‘excess migration’ from Syria's
northeast during 2008e09: a proportion of the estimated
40e60,000 families (i.e. 200e500,000). These figures would suggest
that ‘excess migration’ from northeast Syria during 2008e09
accounted for between 4 and 12% of Syria's 2003e10 urban growth.
We wish to stress that this is only a only rough calculation intended
to illustrate the extent to which Kelley et al. overstate the contribution
of drought to Syria's pre-civil war population pressures.
Indeed, if anything, the above calculation may overstate the
contribution of the 2006/07e2008/09 drought to Syria's urban
growth, for four reasons. First, the 40e60,000 families figure is an
estimate of absolute migration numbers during 2008e09, not an
estimate of ‘excess migration’ above and beyond pre-2008 migration
levels; whatever the correct absolute number, excess migration
must have been considerably lower. Second, drought was not
the only catalyst of ‘excess migration’ from northeast Syria during
2008e09, as discussed in the main text. Third, not all of the ‘excess
migration’ would have contributed to Syria's urban growth given
that many of those leaving northeast Syria during 2008e09
migrated overseas (there was a noticeable rise in Syrian Kurdish
migration to Europe during these years: ACCORD and DIS, 2010).
And third, not all of the internal migration from northeast Syria
during 2008e09 was to urban or peri-urban areas; some of it was
rural-to-rural migration, as discussed in the main text.
Interview method
Interviews with Syrian refugees in Jordan were undertaken by
Fr€ohlich between September 2014 and January 2015. 32 interviews
were undertaken in total (27 in Zaatari camp, three in Azraq camp,
one in Irbid and one in Ramtha) lasting a total of 25 hours, 40 minutes.
The interviewees were all Syrians who had worked full or
part time in agriculture in pre-civil war Syria (specifically, in and
around Dara'a, Homs, Idlib, Damascus, Deir ez-Zor and Raqqa); they
included both male and female respondents. The interviews were
all semi-structured. Some of them were with a single respondent,
while others took the form of focus group discussions in which
more than one person responded to questions. All interviews were
conducted in the respondents' homes in Arabic, and facilitated by a
translator. In Zaatari the translator was Syrian; translators in Azraq,
Irbid and Ramtha were Jordanian nationals.
Interviews focused mainly on respondents' experiences prior to
the civil war, though some questions also addressed their experiences
as refugees. Interviews addressed five areas: (1) interviewees'
general profile, e.g. the size of their household, the
extent of their agricultural income or land ownership; (2) their
experiences of changes in the availability in agricultural inputs, e.g.
water, seeds, and their impacts; (3) their experiences of internal
migration; (4) their political involvement and attitudes; and (5)
their experiences of being a refugee in Jordan.
Respondents were selected by dimensional sampling. The
following characteristics were taken into account in building the
sample (with everyone meeting the basic criteria of Syrian nationality
and having worked in agriculture in Syria pre-2011):
gender, geographical origin, status (migrant/non-migrant), protest
participant/non-participant, landholder/non-landholder. Interviews
were generally arranged with the help of an independent
‘fixer’ rather than through the UNHCR system, in order to avoid
potential biases. Azraq was an exception: here Fr€ohlich was
J. Selby et al. / Political Geography 60 (2017) 232e244242
accompanied by an UNHCR official at all times and, even though
this person never actively interacted with respondents, this may
have affected responses.
References
Ababsa, Myriam (2011). Agrarian counter-reform in Syria (2000-2010) In Raymond
Hinnebusch et al, Agriculture and Reform in Syria (pp. 83e107). Fife: St Andrews
Centre for Syrian Studies.
Ababsa, Myriam (2015). The end of a world: Drought and agrarian transformation in
northeast Syria (2007e2010). In Raymond Hinnebusch, & Tina Zintl (Eds.),
Political economy and international relations: Vol. 1. Syria from reform to revolt
(pp. 199e222). Syracuse: Syracuse University Press.
ACCORD and DIS. (2010). Human rights issues concerning Kurds in Syria. Vienna and
Copenhagen: Austrian Centre for Country of Origin and Asylum Research and
Documentation, and Danish Immigration Service, May. available at: https://
www.nyidanmark.dk/NR/rdonlyres/FF03AB63-10A5-4467-A038-
20FE46B74CE8/0/Syrienrapport2010pdf.pdf (Accessed 15 March 2017).
Adelphi, et al. (2015). A new climate for peace: Taking action on climate and fragility
risks. independent report commissioned by the G7; available at: https://www.
newclimateforpeace.org/#report-top (Accessed 22 December 2015).
Adger, Neil (2006). Vulnerability. Global Environmental Change, 16(3), 268e281.
Aita, Samir (2009a). Labour markets performance and migration flows in Syria. national
background paper. Robert Schuman Centre for Advanced Studies.
Aita, Samir (2009b). Labour markets policies and institutions, with a focus on inclusion,
equal opportunity and the informal economy: The case of Syria. national
background paper. International Labour Organization.
al-Hindi, Atieh (2011). Syria’s agricultural sector: situation, role, challenges and
prospects, In Raymond Hinnebusch et al, Agriculture and Reform in Syria (pp.
15e55). Fife: University of St Andrew Centre for Syrian Studies.
Baker, Aryn (2015). How climate change is behind the surge of migrants to Europe.
Time Magazine (7 September); available at: http://time.com/4024210/climatechange-migrants/
(Accessed 22 December 2015).
Barout, Muhammad Jamal (2011). Syria in the last decade: The dialectic of stagnation
and reform (in Arabic). Doha: Arab Center for Policy and Research Studies.
Becker, Andreas, et al. (2013). A description of the global land-surface precipitation
data products of the Global Precipitation Climatology Centre with sample applications
including centennial (trend) analysis from 1901epresent. Earth Systems
Science Data, 5(1), 71e99.
Bellprat, Omar, & Doblas-Reyes, Francisco (2016). Attribution of extreme weather
and climate events overestimated by unreliable climate simulations. Geophysical
Research Letters, 43(5), 2158e2164.
Bennett, Craig (2015). Failure to act on climate change means an even bigger refugee
crisis. The Guardian (7 September); available at: http://www.theguardian.com/
environment/2015/sep/07/climate-change-global-warming-refugee-crisis
(accessed 22 December 2015).
Black, Richard, et al. (2011). The effect of environmental change on human migration.
Global Environmental Change, 21(1), S3eS11.
Box, Jason, & Klein, Naomi (2015). Why a climate deal is the best hope for peace. The
New Yorker (18 November); available at: http://www.newyorker.com/news/
news-desk/why-a-climate-deal-is-the-best-hope-for-peace (Accessed 22
March 2016).
Brand, Russell (2015). ISIS versus climate change e which kills more? The Trews (5
March); available at: http://www.russellbrand.com/videos/isis-vs-climatechange-which-kills-more-the-trews-e270/
(Accessed 22 December 2015).
Buhaug, Halvard (2010). Climate not to blame for African civil wars. Proceedings of
the National Academy of Sciences, 107(38), 16477e16478.
Buhaug, Halvard, & Urdal, Henrik (2013). An urbanization bomb? Population
growth and social disorder in cities. Global Environmental Change, 23(1), 1e10.
Chalcraft, John (2009). The invisible cage: Syrian migrant workers in Lebanon. Stanford:
Stanford University Press.
Channel 4. (2015). Interview with Ian Bremmer. Channel 4 News (16 June); available
at: https://www.facebook.com/Channel4News/videos/10153008539311939/
(Accessed 22 December 2015).
CNA Military Advisory Board. (2007). National security and the threat of climate
change. Alexandria, VA: Center for Naval Analysis Corporation. available at:
https://www.cna.org/mab/reports (Accessed 22 December 2015).
CNA Military Advisory Board. (2014). National security and the accelerating risks of
climate change. Alexandria, VA: Center for Naval Analysis Corporation. available
at: https://www.cna.org/mab/reports (Accessed 22 December 2015).
Cole, Juan (2015). Did ISIL arise partly because of climate change? The Nation (24 July
24); available at: http://www.thenation.com/article/did-isil-arise-partlybecause-of-climate-change/
(Accessed 22 December 2015).
De Ch^atel, Francesca (2014). The role of drought and climate change in the Syrian
uprising: Untangling the triggers of the revolution. Middle Eastern Studies, 50(4),
521e535.
De Schutter, Olivier (2010). “Two to three million Syrians face food insecurity,” according
to UN food expert. press release, 7 September; available at: http://www.
srfood.org/en/mission-to-syria-four-years-of-drought (Accessed 19 July 2016).
De Schutter, Olivier (2011). Report of the special Rapporteur on the right to food,
Olivier De Schutter, mission to the Syrian Arab Republic. report to UN Human
Rights Council, 27 January (doc: A/HRC/16/49/Add.2); available at: http://www.
srfood.org/en/official-reports (Accessed 15 July 2016).
Democracy Now!. (2015). Martin O'Malley: Climate change helped spark destabilization
of Syria, rise of Islamic state (10 September); available at: http://www.
democracynow.org/2015/9/10/martin_omalley_climate_change_helped_spark
(Accessed 22 December 2015).
Dinshaw, Fram (2015a). Climate change opened “gates of hell” in Syria: Al Gore. National
Observer (9 July); available at: http://www.nationalobserver.com/2015/
07/09/news/climate-change-opened-gates-hell-syria-al-gore (Accessed 22
December 2015).
Dinshaw, Fram (2015b). This is what a climate refugee looks like. National Observer (4
September); available at: http://www.nationalobserver.com/2015/09/04/news/
what-climate-refugee-looks (Accessed 22 December 2015).
Dong, Buwen, & Sutton, Rowan (2015). Dominant role of greenhouse-gas forcing in
the recovery of Sahel rainfall. Nature Climate Change, 5(8), 757e760.
El Laithy, Heba, & Abu-Ismael, Khalid (2005). Poverty in Syria, 1996e2004: Diagnosis
and pro-poor policy considerations. Damascus: United Nations Development
Programme. available at: https://www.google.co.uk/search?
q¼þUNDPþPovertyþinþSyriaþ2005&ie¼utf-8&oe¼utf-8&client¼firefoxb&gfe_rd¼cr&ei¼EIukV_r-I8yN8Qeu77mIDg
(Accessed 5 August 2016).
El Laithy, Heba, & Abu-Ismael, Khalid (2009). Poverty and distribution in Syria.
Damascus: United Nations Development Programme.
Femia, Francesco, & Werrell, Caitlin (2012). Syria: Climate change, drought and social
unrest. Briefer no. 11. Washington, DC: Center for Climate and Security (29
February); available at: http://climateandsecurity.org/reports/ (Accessed 22
December 2015).
Field, Christopher, et al. (Eds.). (2014). Climate change 2014: Impacts, adaptation, and
vulnerability. Part A: Global and sectoral aspects. Contribution of working group II
to the fifth assessment report of the intergovernmental panel on climate change.
Cambridge: Cambridge University Press.
Friedman, Thomas (2012). The other Arab spring. The New York Times (7 April);
available at: http://www.nytimes.com/2012/04/08/opinion/sunday/friedmanthe-other-arab-spring.html
(Accessed 22 December 2015).
Friedman, Thomas (2013). Without water, revolution. The New York Times (18 May);
available at: http://www.nytimes.com/2013/05/19/opinion/sunday/friedmanwithout-water-revolution.html
(Accessed 22 December 2015).
Fr€ohlich, Christiane (2016). Climate migrants as protestors? Dispelling misconceptions
about global environmental change in pre-revolutionary Syria.
Contemporary Levant, 1(1), 38e50.
Gleditsch, Nils Petter, & Nordås, Ragnhild (2014). Conflicting messages? The IPCC on
conflict and human security. Political Geography, 43, 82e90.
Gleick, Peter (2014). Water, drought, climate change, and conflict in Syria. Weather,
Climate and Society, 6(3), 331e340.
Goldstone, Jack (2002). Population and security: How demographic change can lead
to violent conflict. Journal of International Affairs, 56(1), 3e22.
Goulden, Robert (2011). Housing, inequality, and economic change in Syria’. British
Journal of Middle Eastern Studies, 38(2), 187e202.
Gray, Richard (2015). Did climate change trigger the war in Syria? Severe drought may
have contributed to uprising, study reveals. Daily Mail (2 March); available at:
http://www.dailymail.co.uk/sciencetech/article-2976063/Did-climate-change-
trigger-war-Syria-Severe-drought-contributed-uprising-study-reveals.html
(Accessed 22 December 2015).
Harris, Ian, et al. (2014). Updated high-resolution grids of monthly climatic observations
- the CRU TS3.10 dataset. International Journal of Climatology, 34(3),
623e642.
Hartmann, Betsy (2010). Rethinking climate refugees and climate conflict: Rhetoric,
reality and the politics of policy discourse. Journal of International Development,
22(2), 233e246.
Hendrix, Cullen S., & Salehyan, Idean (2012). Climate change, rainfall, and social
conflict in Africa. Journal of Peace Research, 49(1), 35e50.
He, Jie, & Soden, Brian (2017). A re-examination of the projected subtropical precipitation
decline. Nature Climate Change, 7(1), 53e57.
Hockenos, Paul (2015). The climate wars are coming e and more refugees with them.
Al Jazeera (17 September); available at: http://america.aljazeera.com/opinions/
2015/9/the-climate-wars-are-comingeand-more-refugees-with-them.html
(Accessed 22 December 2015).
Hoerling, Martin, et al. (2012). On the increased frequency of Mediterranean
drought’. Journal of Climate, 25(6), 2146e2161.
Hole, Frank (2009). Drivers of unsustainable land use in the semi-arid Khabour
River Basin, Syria. Geographical Research, 47(1), 4e14.
Hsiang, Solomon, et al. (2011). Civil conflicts are associated with the global climate.
Nature, 476, 438e440.
Hulme, Mike (1996). Recent climate change in the world's drylands. Geophysical
Research Letters, 23(1), 61e64.
Hulme, Mike (2001). Climatic perspectives on Sahelian desiccation: 1973e1998.
Global Environmental Change, 11(1), 19e29.
Hulme, Mike, et al. (2001). ‘African climate change: 1900-2100’. Climate Research
(Vol. 17,(No. 2), 145e168.
Hulme, Mike (2014). Attributing weather extremes to “climate change”: a review.
Progress in Physical Geography, 38(4), 499e511.
IRIN. (2009a). Drought driving farmers to the cities. Integrated Regional Information
Networks (2 September); available at: http://reliefweb.int/report/syrian-arabrepublic/syria-drought-driving-farmers-cities
(Accessed 20 July 2016).
IRIN. (2009b). Syria: Drought response faces funding shortfall. Integrated Regional
Information Networks (24 November); available at: http://www.irinnews.org/
printreport.aspx?reportid¼87165 (Accessed 22 December 2015).
Jaubert, Ronald, & Bocco, Riccardo (1998). Traditional dryland resource
J. Selby et al. / Political Geography 60 (2017) 232e244 243
management systems in Syria: Construction or reality? In Clarke John, &
Noin Daniel (Eds.), Population and environment in arid regions (pp. 177e192).
Paris: UNESCO.
Juncker, Jean-Claude (2015). State of the union 2015: Time for honesty, unity and
solidarity. Strasbourg (9 September); available at: http://europa.eu/rapid/pressrelease_SPEECH-15-5614_en.htm
(Accessed 22 December 2015).
Kelley, Colin, et al. (2015). Climate change in the fertile crescent and implications of
the recent Syrian drought. Proceedings of the National Academy of Sciences,
112(11), 3241e3246.
Kerry, John (2015). Remarks by the secretary of state at old dominion university on
climate change and national security (10 November); available at: http://www.
state.gov/secretary/remarks/2015/11/249393.htm (Accessed 22 December
2015).
Kevane, Michael, & Gray, Leslie (2008). Darfur: rainfall and conflict. Environmental
Research Letters, 3(3), 1e10.
Khawaja, Marwan (2002). Internal migration in Syria: Findings from a national survey.
Oslo: Fafo.
Ki Moon, Ban (2007). A climate culprit in Darfur. The Washington Post (16 June);
available at: http://www.washingtonpost.com/wp-dyn/content/article/2007/
06/15/AR2007061501857.html (Accessed 22 December 2015).
King, David, et al. (2015). Climate change: A risk assessment, report commissioned by
UK foreign and commonwealth office. Cambridge: Centre for Science and Policy.
available at: http://www.csap.cam.ac.uk/projects/climate-change-riskassessment/
(Accessed 22 December 2015).
Leenders, Reinoud (2012). Collective action and mobilization in Dar'a: An anatomy
of the onset of Syria's popular uprising. Mobilization, 17(4), 419e434.
Leenders, Reinoud, & Heydemann, Steven (2012). Population mobilization in Syria:
Opportunity and threat, and the social networks of the early risers. Mediterranean
Politics, 17(2), 139e159.
Leybourne, Marina (1998). The adaptability of Syrian Bedouin production systems.
In John Clarke, & Daniel Noin (Eds.), Population and environment in arid regions
(pp. 193e207). Paris: UNESCO.
MAAR. (2016). Annual agricultural statistical abstract, 1991e2014. Syrian Ministry of
Agriculture and Agrarian Reform; in Arabic. available at: http://moaar.gov.sy/
main/archives/category/%D8%A7%D9%84%D9%85%D8%AC%D9%85%D9%88%D8%
B9%D8%A7%D8%AA-%D8%A7%D9%84%D8%A7%D8%AD%D8%B5%D8%A7%D8%A6%
D9%8A%D8%A9 (Accessed 12 March 2016).
Malm, Andreas (2016). Revolution in a warming world: Lessons from the Russian to
the Syrian revolutions. In Gregory Albo, & Leo Panitch (Eds.), Socialist register
2017: Rethinking revolutions (pp. 120e142). London: Monthly Review.
Massoud, Ali (2010). Years of drought: A report on the effects of drought on the Syrian
peninsula. Berlin: Heinrich B€oll-Stiftung. available at: https://lb.boell.org/en/
2010/12/14/years-drought-report-effects-drought-syrian-peninsula (Accessed
22 December 2015).
Mazo, Jeffrey (2010). Climate conflict: How global warming threatens security and
what to do about it. London: International Institute for Strategic Studies.
McSweeny, Robert (2015). Updated: The climate change papers most featured in the
media. CarbonBrief (29 July); available at: http://www.carbonbrief.org/updatedthe-climate-change-papers-most-featured-in-the-media/
(Accessed 22
December 2015).
Mills, Rhiannon (2015). Charles: Syria's war linked to climate change. Sky News (23
November); available at: http://news.sky.com/story/1592373/charles-syriaswar-linked-to-climate-change
(Accessed 22 December 2015).
Nordås, Ragnhild, & Gleditsch, Nils Petter (2007). Climate change and conflict. Political
Geography, 26(6), 628e631.
Obama, Barack (2009). A just and lasting peace. Oslo: Nobel Peace Prize Lecture (10
December); available at: http://www.nobelprize.org/nobel_prizes/peace/
laureates/2009/obama-lecture_en.html (Accessed 22 December 2015).
Obama, Barack (2015). Remarks by the president at the United States coast guard
academy commencement (20 May); available at: https://www.whitehouse.gov/
the-press-office/2015/05/20/remarks-president-united-states-coast-guardacademy-commencement
(Accessed 22 December 2015).
O'Hagan, Ellie May (2015). Mass migration is no “crisis”: It's the new normal as the
climate changes. The Guardian (18 August); available at: http://www.
theguardian.com/commentisfree/2015/aug/18/mass-migration-crisis-refugeesclimate-change
(Accessed 22 December 2015).
Quinn, Audrey, & Roche, Jackie (2014). Syria's climate conflict. cartoon for Years of
Living Dangerously platform; available at: http://yearsoflivingdangerously.
tumblr.com/post/86898140738/this-comic-was-produced-in-partnership-byyears-of
(Accessed 10 February 2017).
Reuveny, Rafael (2007). Climate change induced migration and violent conflict.
Political Geography, 26(6), 656e673.
Saleeby, Suzanne (2012). Sowing the seeds of dissent: Economic grievances and the
Syrian social contract's unravelling. Jadaliyya (16 February); available at: http://
www.jadaliyya.com/pages/index/4383/sowing-the-seeds-of-dissent_
economic-grievances-an (Accessed 22 December 2015).
Sanderson, Benjamin, et al. (2015). A representative democracy to reduce interdependency
in a multimodel ensemble. Journal of Climate, 28(13), 5171e5194.
Schulman, Jeremy (2015). Bernie Sanders: Yes, climate change is still our biggest national
security threat. Mother Jones (14 November); available at: http://www.
motherjones.com/environment/2015/11/bernie-sanders-climate-change-isis
(Accessed 22 December 2015).
Selby, Jan (2014). Positivist climate conflict research: a critique. Geopolitics (Vol.
19,(No. 4), 829e856.
Selby, Jan, & Hoffmann, Clemens (2014a). Beyond scarcity: rethinking water, climate
change and conflict in the Sudans. Global Environmental Change, 29, 360e370.
Selby, Jan, & Hoffmann, Clemens (2014b). Rethinking climate change, conflict and
security. Geopolitics, 19(4), 747e756.
SCPR. (2013). Socioeconomics roots and impact of the Syrian crisis. Damascus: Syrian
Center for Policy Research. available at: http://scpr-syria.org/publications/
policy-reports/socioeconomic-roots-and-impact-of-the-syrian-crisis-2013
(Accessed 7 September 2016).
Showtime. (2014). Years of living dangerously, season 1, episode 1 (13 April); available
at: https://vimeo.com/89648161 (Accessed 22 December 2015).
Solow, Andrew (2015). Extreme weather, made by us? Science, 349(6255),
1444e1445.
Sternberg, Troy (2013). Chinese drought, wheat, and the Egyptian uprising: How a
localized hazard became globalized. In Caitlin Werrell, & Francesco Femia (Eds.),
The Arab spring and climate change (pp. 7e14). Washington DC: Center for
American Progress, Stimson Center, and Center for Climate and Security.
available at: https://www.americanprogress.org/issues/security/report/2013/
02/28/54579/the-arab-spring-and-climate-change/ (Accessed 22 December
2015).
Stocker, Thomas et al. (Eds.), (2014). Climate change 2013: The physical science basis.
Contribution of working group I to the fifth assessment report of the intergovernmental
panel on climate change, Cambridge: Cambridge University Press.
Theisen, Ole Magnus, et al. (2011/12). Climate wars? Assessing the claim that
drought breeds conflict. International Security, 36(3), 79e106.
Trigo, Ricardo, et al. (2010). The intense 2007e2009 drought in the fertile crescent:
Impacts and associated atmospheric circulation. Agricultural and Forest Meteorology,
150(9), 1245e1257.
UN-OCHA. (2008). Syria drought appeal, September 2008. New York: UN Office for
the Coordination of Humanitarian Affairs, 29 September. available at: http://
www.unocha.org/cap/appeals/syria-drought-appeal-2008 (Accessed 31 March
2016).
UN-OCHA. (2009a). Syria drought response plan. New York: UN Office for the Coordination
of Humanitarian Affairs, 11 August. available at: http://www.unocha.
org/cap/appeals/drought-response-plan-syria-2009 (Accessed 31 March 2016).
UN-OCHA. (2009b). Joint united nations drought assessment mission, the Syrian Arab
Republic 2008/09. New York: UN Office for the Coordination of Humanitarian
Affairs, 9 September.
UN-OCHA. (2010). Syria drought response plan 2009e10, mid-term review. New York:
UN Office for the Coordination of Humanitarian Affairs, February. available at:
http://reliefweb.int/sites/reliefweb.int/files/resources/
20E00ADAF9F3C153852576D20068E86B-Full_Report.pdf (Accessed 31 March
2016).
UNICEF. (2014). Multidimensional poverty in Syria. Damascus: UNICEF Country Office.
available at: http://www.cbssyr.sy/Child_Poverty_Multidimensional/
Multidimensional_Poverty_in_Syria_EN.pdf (Accessed 15 March 2017).
Urdal, Henrik (2005). People versus Malthus: Population pressure, environmental
degradation, and armed conflict revisited. Journal of Peace Research, 42(4),
417e434.
US Embassy Damascus. (2008a). Political impact of Syria's records-low wheat harvest.
US Diplomatic Cable 08DAMASCUS518_a (20 July); available at: https://
wikileaks.org/plusd/cables/08DAMASCUS518_a.html (Accessed 8 September
2016).
US Embassy Damascus. (2008b). 2008 UN drought appeal for Syria. US Diplomatic
Cable 08DAMASCUS847_a (26 November); available at: https://wikileaks.org/
plusd/cables/08DAMASCUS847_a.html (accessed 8 September 2016).
US Embassy Damascus. (2008c). Syrians adjusting to “new normal” of inflation. US
Diplomatic Cable 08DAMASCUS587_a (18 August); available at: https://
wikileaks.org/plusd/cables/08DAMASCUS587_a.html (Accessed 24 October
2016).
Verhoeven, Harry (2011). Climate change, conflict and development in Sudan:
Global neo-Malthusian narratives and local power struggles. Development and
Change, 42(3), 679e707.
Verme, Paolo, et al. (2016). The welfare of Syrian refugees: Evidence from Jordan and
Lebanon. Washington DC: World Bank.
Welzer, Harald (2012). Climate wars: What people will be killed for in the 21st century.
Cambridge: Polity Press.
Werrell, Caitlin, et al. (2015). Did we see it coming? State fragility, climate
vulnerability, and the uprisings in Syria and Egypt. SAIS Review, 35(1), 29e46.
Werrell, Caitlin, & Femia, Francesco (2015). Let's not say climate change causes war.
But let's not also ignore the real security risks. Center for Climate and Security
Blog (5 December); available at: http://climateandsecurity.org/2015/12/05/lets-
not-say-climate-change-causes-war-but-lets-also-not-ignore-the-real-risks/
(Accessed 22 December 2015).
Westlake, Michael (2003). The economics of strategic crops. In Ciro Firilllo, &
Jacques Vercueil (Eds.), Syrian agriculture at the crossroads (pp. 137e162). Rome:
UN Food and Agriculture Organization.
World Bank. (2016). Agriculture, value added (% of GDP). World Bank DataBank.
available at: http://data.worldbank.org/indicator/NV.AGR.TOTL.ZS (Accessed 19
September 2016).
Zastrow, Mark (2015). Climate change implicated in current Syrian conflict. Nature
(2 March); available at: http://www.nature.com/news/climate-changeimplicated-in-current-syrian-conflict-1.17027
(Accessed 22 December 2015).
J. Selby et al. / Political Geography 60 (2017) 232e244244