Subnational violent conflict forecasts for sub-Saharan Africa, 2015–65, using climate-sensitive models
| Author | Andrew M Linke,Frank DW Witmer,John O’Loughlin,Arlene Laing,Andrew Gettelman |
| Published date | 01 March 2017 |
| Date | 01 March 2017 |
| DOI | http://doi.org/10.1177/0022343316682064 |
| Subject Matter | Research Articles |
Subnational violent conflict forecasts
for sub-Saharan Africa, 2015–65,
using climate-sensitive models
Frank DW Witmer
Department of Computer Science & Engineering, University of Alaska Anchorage
Andrew M Linke
Department of Geography, University of Utah
John O’Loughlin
Institute of Behavioral Science and Department of Geography, University of Colorado Boulder
Andrew Gettelman
National Center for Atmospheric Research, Boulder, CO
Arlene Laing
Cooperative Institute for Research in the Atmosphere, Colorado State University
Abstract
How will local violent conflict patterns in sub-Saharan Africa evolve until the middle of the 21st century? Africa is
recognized as a particularly vulnerable continent to environmental and climate change since a large portion of its
population is poor and reliant on rain-fed agriculture. We use a climate-sensitive approach to model sub-Saharan African
violence in the past (geolocated to the nearest settlements) and then forecast future violence using sociopolitical factors such
as population size and political rights (governance), coupled with temperature anomalies. Our baseline model is calibrated
using 1gridded monthly data from 1980 to 2012 at a finer spatio-temporal resolution than existing conflict forecasts. We
present multiple forecasts of violence under alternative climate change scenarios (optimistic and current global trajectories),
of political rights scenarios (improvement and decline), and population projections (low and high fertility). We evaluate
alternate shared socio-economic pathways (SSPs) by plotting violence forecasts over time and by detailed mapping
of recent and future levels of violence by decade. The forecasts indicate that a growing population and rising temperatures
will lead to higher levels of violence in sub-Saharan Africa if political rights do not improve. If political rights continue
to improve at the same rate as observed over the last three decades, there is reason for optimism that overall levels of
violence will hold steady or even decline in Africa, in spite of projected population increases and rising temperatures.
Keywords
disaggregated violence, environmental change, governance, multilevel models, population projections,
socioeconomic pathways
Introduction
How will the local patterns of violent conflict in sub-
Saharan Africa change through the middle of the 21st
century? Africa is recognized as a particularly vulnerable
Corresponding author:
fwitmer@alaska.edu
Journal of Peace Research
2017, Vol. 54(2) 175–192
ªThe Author(s) 2017
Reprints and permission:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0022343316682064
journals.sagepub.com/home/jpr
continent to environmental and climate change since a
large portion of its population is poor and is reliant on
rain-fed agriculture. Our research uses a climate-sensitive
approach to model sub-Saharan African violence in the
past and then forecasts violence trends using key expla-
natory variables that have been shown to influence con-
flict. Forecasting the future is an exercise fraught with
uncertainty but valuable when used as a tool to explore
the outcomes of plausible political, social and climate
scenarios. There is a growing literature on scenario devel-
opment in the context of environmental decision making
(Mahmoud et al., 2009; Steinitz et al., 2003) and climate
change (O’Neill et al., 2014), and we adopt this general
approach to explore shared socio-economic pathways
(SSPs) for violent conflict in sub-Saharan Africa. In par-
ticular, we consider future scenarios that vary fertility
assumptions, governance trends and climate projections.
Conflict forecasting in the study of
environmental change and conflict
Predictions are increasingly used in conflict research and
are based on two general motivations. One approach uses
predictions (either in- or out-of-sample validation) to
assess the influences of independent variables upon an
outcome of interest (Ward, Greenhill & Bakke, 2010;
O’Loughlin, Linke & Witmer, 2014; O’Loughlin et al.,
2012; Wischnath & Buhaug, 2014). Another uses mod-
eling and simulation techniques to forecast observed
trends into the future. Interest in this second kind of
analysis has been growing in recent years (Hegre et al.,
2013, 2016, 2017; Ward et al., 2013; Blair, Blattman &
Hartman, 2017; Beger, Dorff & Ward, 2014; O’Brien,
2010; Schneider, Gleditsch & Carey, 2011; Schrodt,
Yonamine & Bagozzi, 2013). We are primarily moti-
vated by this second type of forecasting, though initial
steps in our research rely on model validation using
observed data for recent years.
We forecast conflict using multiple future climate
scenarios to address a question of great interest to aca -
demic and policy communities: how do climate varia-
bility and environmental stress lead to violent conflict?
In doing so, we are contributing to a growing body of
literature published in Science (Hsiang, Burke &
Miguel, 2013), Nature (Hsiang, Meng & Cane,
2011), Proceedings of the National Academy of Sciences
(Buhaug, 2010; Burke et al., 2009; O’Loughlin, Linke
& Witmer, 2014; O’Loughlin et al., 2012; Schleussner
et al., 2016) and Global Environmental Change (Bo
¨h-
melt et al., 2014; Ide et al., 2014; Linke et al., 2015).
Similarly, journals in the fields of political geography
(De Juan, 2015), development economics (Maystadt,
Calderone & You, 2015), clima te studies (Theisen, Gle-
ditsch & Buhaug, 2013), and peace research (Koubi
et al., 2012) contain valuable contributions that con-
sider particular social conditions that enable climatolo-
gical extreme events to affect societal conflict. Robust
debate continues about the relative importance of cli-
mate factors on conflict with Buhaug et al. (2014) argu-
ing that Hsiang, Burke & Miguel (2013) over reach in
their conclusion that deviations from normal tempera-
tures increase the risk of conflict across multiple tem-
poral and spatial scales.
Our contribution to the climate–conflict body of
research therefore moves beyond the study of only the
past and present. To that end, we consider two specific
questions in the context of the climate–conflict litera-
ture. First, what will future patterns of conflict look like?
Second, what factors are driving those future patterns?
None of the research cited above includes substantial
engagement with projected climate scenarios though
generalized statements have been made. Hsiang, Burke
& Miguel (2013: 1235367) argue, for example, ‘because
locations throughout the inhabited world are expected to
warm two to four standard deviations by 2050, amplified
rates of human conflict could represent a large and crit-
ical impact of anthropogenic climate change’. But where
would we observe such a spike in violence and under
what circumstances? Answering these questions is a logi-
cal extension of existing research and is a key goal in this
study. We quantify expectations about violent conflict
patterns in geographically disaggregated predictions for
sub-Saharan Africa until 2065.
In modeling and forecasting violence across all of sub-
Saharan Africa at a 1(degree) grid resolution, we expli-
citly consider the effects of temperature variability, while
controlling for temporal reporting bias in the coding of
our violent event data. Additionally, we account for a
number of key social and political variables that have
known associations with violence. The incorporation
of these effects into regression analyses has been debated;
while some encourage a thorough and dedicated effort to
capture the specific geographical contexts and social set-
tings of violence (e.g. Raleigh, Linke & O’Loughlin,
2014), others do not explicitly include these possible
influences in statistical estimations even though data are
available (e.g. Hsiang, Burke & Miguel, 2013). By
including these sociopolitical factors with temperature
variability, we are able to evaluate the relative contribu-
tion of each.
In this article, we consider especially the role of polit-
ical settings within which conflicts emerge. There is
176 journal of PEACE RESEARCH 54(2)
To continue reading
Request your trial