Civil war, climate change, and development: A scenario study for sub-Saharan Africa
| Date | 01 January 2012 |
| Published date | 01 January 2012 |
| Author | Richard SJ Tol,Conor Devitt |
| DOI | 10.1177/0022343311427417 |
| Subject Matter | Research Articles |
Civil war, climate change, and
development: A scenario study
for sub-Saharan Africa
Conor Devitt
Booth School of Business, University of Chicago
Richard SJ Tol
Economic and Social Research Institute, Dublin, Vrije Universiteit, Amsterdam & Trinity College, Dublin
Abstract
This article presents a model of development, civil war and climate change. There are multiple interactions.
Economic growth reduces the probability of civil war and the vulnerability to climate change. Climate change
increases the probability of civil war. The impacts of climate change, civil war and civil war in the neighbouring
countries reduce economic growth. The model has two potential poverty traps – one is climate-change-induced and
one is civil-war-induced – and the two poverty traps may reinforce one another. The model is calibrated to sub-
Saharan Africa and a double Monte Carlo analysis is conducted in order to account for both parameter uncertainty
and stochasticity. Although the IPCC Special Report on Emission Scenarios (SRES) is used as the baseline, thus
assuming rapid economic growth in Africa and convergence of African living standards to the rest of the world, the
impacts of civil war and climate change (ignored in SRES) are sufficiently strong to keep a number of countries in
Africa in deep poverty with a high probability.
Keywords
civil war, climate change, economic development
Introduction
The socio-economic scenarios that underpin future
projections of climate change are very peaceful (Nakice-
novic & Swart, 2001). This is in sharp contrast to the
past, which regularly saw violent conflict between and
within states. The absence of (civil) war in future scenar-
ios of climate change is even more surprising when one
considers that violent conflict can have a profound
impact on development (Butkiewicz & Yanikkaya,
2005), and that one of the more worrying assertions is
that climate change could enhance violent conflict
(Barnett & Adger, 2007). This article seeks to fill this
void by developing a simulation model for the three-
way interaction between civil war, climate change and
development.
The model has a few, simple components: climate
change has a negative impact on the economy, slowing
down its growth. Climate change increases the probabil-
ity of civil war. Civil war has a negative impact on eco-
nomic growth. In turn, economic growth reduces the
vulnerability to the impacts of climate change, and it
reduces the probability of an outbreak of violent conflict.
Although its components are simple, when put together
the model is complex.
As far as we know, this is the first attempt to study the
three-way interaction between climate change, civil war,
and development. Essentially, we model a race.
Corresponding author:
richard.tol@esri.ie
Journal of Peace Research
49(1) 129–145
ªThe Author(s) 2012
Reprints and permission:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0022343311427417
jpr.sagepub.com
p
eace
R
ESEARCH
journal of
Economic growth reduces the risk of conflict and the
impact of climate change. But climate change and con-
flict reinforce one another and reduce economic growth.
If the first effect is stronger, countries will be rich, peace-
ful and not much bothered by climate change. If the lat-
ter effect is stronger, countries will be poor, torn by
conflict and suffer from climate change. Phrased like
this, the model is used to investigate whether there is a
conflict-and-climate-induced poverty trap – or rather,
the size of the trap and which countries are more likely
to be caught by it.
We qualitatively sketch the mechanisms above. It is
therefore possible to construct a mathematical model
from which a conflict-and-climate poverty trap emerges.
We do so below. We parameterize the model with realis-
tic values and conduct a systematic sensitivity analysis on
the parameters. This exercise takes the article from the
question ‘is it possible?’ to ‘how likely is it?’.
It would be preferable to investigate the strength of
the hypothesized conflict-and-climate poverty trap using
observations. However, rapid climate change has not
happened in the period for which there are good data
on conflict and development. We therefore rely on a
simulation model.
While there are a number of articles on the relation-
ship between conflict and economic growth and on cli-
mate change and growth, there is little quantitative
evidence on conflict and climate change – see the next
section for a literature review. Therefore, as a secondary
contribution, the article also develops and estimates a
model of the impact of climate change on civil war.
As a third contribution, we introduce a new richness
to the scenarios of development used in climate change
analysis. We apply the model to sub-Saharan Africa, the
region that is least developed and most subject to (civil)
war.
The article proceeds as follows. The next section
reviews the literature. After that, we present the model,
with additional material in the appendix. This is fol-
lowed by a discussion of the results. As with any numer-
ical model, the results follow from the assumptions,
which are particularly uncertain in this case. We there-
fore conduct a systematic sensitivity analysis and focus
on the qualitative results. The final section concludes.
Previous literature
Climate and conflict
Existing empirical research on the role of climate change
in violent conflict is limited and inconclusive. Homer-
Dixon (1994) examines a number of case studies, in
order to determine if environmental scarcities cause
violent conflict. Evidence from these case studies sug-
gests that while conflict has indeed occurred in areas of
resource scarcity, key contextual factors have played an
important role. For example, he argues that serious civil
unrest is unlikely to occur unless the political structure
prevents challenger groups from expressing their grie-
vances peacefully, but offers these groups an opportunity
for violence against authority. Later research (Buhaug,
2010a,b; Buhaug & Rød, 2006; Burke et al., 2009,
2010; Dixon, 2009; Gleditsch, 1998; Gleditsch et al.,
2006; Hauge & Ellingsen, 1998; Henderson, 2000;
Henderson & Singer, 2000; Hendrix & Glaser, 2007;
Nordås & Gleditsch, 2007; Raleigh, 2010; Raleigh &
Urdal, 2007; Theisen, 2008; Urdal, 2005) finds conflict-
ing evidence about whether or not environment and cli-
mate factors contribute to violent conflict. There is a
consensus, however, that other, non-environmental fac-
tors dominate.
See also the other articles in this special issue, three of
which are particularly relevant for our contribution.
Gartzke (2012) seeks to estimate the effect of the annual
global mean temperature on interstate conflict in the last
150 years but, since he fails to account for non-
stationarity in the data (Engle & Granger, 1987), his
results are not robust. Using panel data with 30 years and
170 countries, Bergholt & Lujala (2012) find that natu-
ral disasters negatively affect economic growth, but that
this does not in turn influence the onset of civil war. In a
similar analysis, Koubi et al. (2012) find that tempera-
ture and precipitation do not affect economic growth,
but that growth does reduce the probability of armed
civil conflict.
Collier & Hoeffler (1998) were the first to suggest an
‘economic theory’ of civil conflict – rent-seeking by vio-
lence – and to test their predictions with data. Later arti-
cles have refined the hypotheses and econometrics
(Brunnschweiler & Bulte, 2009; Collier, Hoeffler &
Rohner, 2009; Collier & Hoeffler, 2005; Elbadawi &
Sambanis, 2000; Justino, 2009; Schollaert & van de
Gaer, 2009; van der Ploeg & Poelhekke, 2010; Welsch,
2008; Wick, 2008; Wick & Bulte, 2006). While these
articles tend to find a link between material deprivation
and conflict and between specific resources and conflict,
there is no direct link between climate and conflict.
Material deprivation has many causes and climate is at
best a contributing factor (Acemoglu, Johnson & Robin-
son, 2001, 2002; Easterly & Levine, 2003; Gallup,
Sachs & Mellinger, 1999; Masters & McMillan,
2001). According to this strand of literature, people may
fight over resources that are highly valuable and easy to
130 journal of PEACE RESEARCH 49(1)
To continue reading
Request your trial