Causal mechanisms in civil war mediation: Evidence from Syria
| DOI | 10.1177/1354066119856084 |
| Date | 01 March 2020 |
| Published date | 01 March 2020 |
| Author | Magnus Lundgren |
research-article2019
EJ R
I
Article
European Journal of
International Relations
Causal mechanisms in
2020, Vol. 26(1) 209 –235
© The Author(s) 2019
civil war mediation:
Article reuse guidelines:
sagepub.com/journals-permissions
https://doi.org/10.1177/1354066119856084
DOI: 10.1177/1354066119856084
Evidence from Syria
journals.sagepub.com/home/ejt
Magnus Lundgren
Stockholm University, Sweden
Abstract
Studies of conflict management by international organizations have demonstrated
correlations between institutional characteristics and outcomes, but questions remain
as to whether these correlations have causal properties. To examine how institutional
characteristics condition the nature of international organization interventions, I examine
mediation and ceasefire monitoring by the Arab League and the United Nations during
the first phase of the Syrian civil war (2011–2012). Using micro-evidence sourced from
unique interview material, day-to-day fatality statistics, and international organization
documentation, I detail causal pathways from organizational characteristics, via intervention
strategies, to intervention outcomes. I find that both international organizations relied
on comparable intervention strategies. While mediating, they counseled on the costs
of conflict, provided coordination points, and managed the bargaining context so as to
sideline spoilers and generate leverage. While monitoring, they verified violent events,
engaged in reassurance patrols, and brokered local truces. The execution of these
strategies was conditioned on organizational capabilities and member state preferences
in ways that help explain both variation in short-term conflict abatement and the long-
term failure of both international organizations. In contrast to the Arab League, the
United Nations intervention, supported by more expansive resources and expertise,
temporarily shifted conflict parties away from a violent equilibrium. Both organizations
ultimately failed as disunity among international organization member state principals cut
interventions short and reduced the credibility of international organization mediators.
Keywords
Bargaining, causal mechanisms, ceasefire, civil war, international organizations,
mediation, Syria
Corresponding author:
Magnus Lundgren, Department of Political Science, Stockholm University, 10691, Sweden.
Email: magnus.lundgren@statsvet.su.se
210
European Journal of International Relations 26(1)
Introduction
Despite increased reliance on international organizations (IOs) in civil war mediation,
the conditions of IO intervention effectiveness remain poorly understood. Claims that
IO mediation shortens civil wars (Gartner, 2011) confront claims that it lacks discerni-
ble effects (Bercovitch and Schneider, 2000; Regan, 2002) or produces fragile settle-
ments prone to disintegration (Beardsley, 2008). One reason for such divergence is the
premise — common to much theoretic and large-N work — that IOs can be studied as
uniform actors with uniform effects. Such “billiard ball” assumptions are mismatched
to a reality where IOs display significant institutional diversity (Hawkins et al., 2006;
Koremenos et al., 2001), disguising factors that may explain why IOs sometimes pro-
mote peace and sometimes not, and why some IOs are more effective than others.
Following calls for the disaggregation of IOs, scholars have begun to examine IO medi-
ation in ways more sensitive to institutional heterogeneity (e.g. Boehmer et al., 2004;
Haftel 2007; Hansen et al., 2008; Tir and Karreth, 2018). However, while this literature
has established correlations between IO characteristics and outcomes, it has failed to
provide credible evidence on underlying mechanisms, leaving the causal nature of insti-
tutional correlations in doubt.1 How does variation in institutional characteristics affect
the nature and outcome of IO civil war interventions?
In this article, I seek to elucidate the causal mechanisms involved in IO conflict man-
agement. In the first part, I theorize the causal mechanisms that link institutional charac-
teristics, via conflict management strategies, to intervention outcomes. Building on the
literatures on bargaining and war, conflict management, and international institutional
design, I conceptualize peace-brokering IOs as consisting of supranational agents
endowed with conflict management capabilities, and member state principals endowed
with preferences over intervention policies. I propose that variation in these institutional
characteristics determines the nature of IO conflict management interventions, via spe-
cific causal mechanisms, with implications for outcomes.
I evaluate the theoretical framework against data on mediation and ceasefire monitor-
ing by the League of Arab States (LAS), more commonly known as the Arab League, and
the United Nations (UN) during the first phase of the civil war in Syria (2011–2012). The
two interventions took place under comparable conditions, providing a rare opportunity
for relatively controlled cross- and within-case inference. While neither intervention
resulted in lasting conflict resolution, variation in short-term outcomes makes meaning-
ful analysis of intervention mechanisms possible. The analysis draws on independently
sourced interview data, time-series analysis of fatality statistics, and leaked and public
IO documentation.
I find that case evidence confirms the general validity of the theoretical framework
but, more importantly, points to mechanisms that explain how institutional characteris-
tics matter in IO civil war interventions. Case evidence demonstrates that the scope and
proficiency of mediation sub-strategies — counseling on the costs of conflict, the provi-
sion of coordination points, and the strategic management of the bargaining context — is
shaped by the extent of tailored institutional capabilities available in IO secretariats and
field locations. Drawing on greater informational capabilities, the UN could maintain a
greater scope and proficiency across all three mediation sub-strategies. It consulted more
widely, designed a ceasefire that better reflected the nature of the conflict, and sought to
Lundgren
211
sideline potential spoilers, including Islamist hardliners and the expatriate opposition.
With lower institutional capabilities and expertise, LAS mediation was constrained to
regional engagements, produced a flawed ceasefire proposal, and did not structure its
mediation process to sideline spoilers or reduce rebel fragmentation.
Similarly, case evidence demonstrates that the extent and quality of IO monitoring
sub-strategies — the verification of violent events, reassurance patrols, and localized
truce brokering — depend on in-house institutional resources. Again, due to its higher
capabilities, the scope and proficiency of the UN monitoring intervention superseded
that of LAS. UN observers covered a greater number of conflict spots and established
mechanisms for public reporting on disputant behavior so that limited tit-for-tat coopera-
tion between combatants could arise. Lacking institutional field mission support, LAS
deployed a monitoring force that was undertrained, under-equipped, and characterized
by ad hoc solutions, undercutting its credibility as an external enforcement facilitator.
The evidence suggests that the UN’s intervention had greater impact on short-term
outcomes, especially in terms of conflict abatement, but neither organization managed to
bring the combatants — the Syrian government and a loose, Sunni-dominated rebel coa-
lition — to a long-term settlement. The war continued despite their attempts to resolve it.
I find that another institutional characteristic — member state preferences — provides an
important part of the explanation for why the two interventions ultimately failed. In both
cases, preference dissonance among IO member states cut interventions short and
reduced the efficacy of the strategies pursued by supranational agents on the ground.
By introducing evidence on the linkages between institutional characteristics, inter-
vention strategies, and outcomes, this study makes several contributions to the literature
on IO conflict management, as further developed in the conclusion. First, it demonstrates
that previously identified correlations between institutional capabilities and outcomes
(Boehmer et al., 2004; Hansen et al., 2008; Lundgren, 2017; Tir and Karreth, 2018) have
causal properties. It provides detailed evidence on the causal mechanisms of IO media-
tion and monitoring, hitherto assumed in general terms but not understood in detail, link-
ing them to variation in institutional characteristics. Second, case evidence suggests that
mediation strategy, frequently identified as a determinant of mediation outcomes (Greig
and Diehl, 2012), is shaped not only by institutional capabilities, but also by IO member-
ship preferences. Third, the evidence suggests that preference shifts among IO principals
are a possible explanation for the “time-inconsistency problems” identified by mediation
scholars (Beardsley, 2008) as a key reason for mediation failure.
In a wider light, the evidence suggests that scholarship on conflict management needs
to move beyond the assumptions of IO uniformity now common in the literature (e.g.
Gartner, 2011; Svensson, 2007) and take institutional heterogeneity more seriously...
To continue reading
Request your trial