International reputation and alliance portfolios: How unreliability affects the structure and composition of alliance treaties
| Author | Neil Narang,Brad L LeVeck |
| Date | 01 May 2019 |
| DOI | 10.1177/0022343318808844 |
| Published date | 01 May 2019 |
| Subject Matter | Regular Articles |
International reputation and alliance
portfolios: How unreliability affects
the structure and composition
of alliance treaties
Neil Narang
Department of Political Science, University of California, Santa Barbara
Brad L LeVeck
Department of Political Science, University of California, Merced
Abstract
Why do states ever form military alliances with unreliable partners? States sign offensive and defensive military
alliances to increase their fighting capabilities in the event of war and as a signal to deter potential aggressors from
initiating a crisis. Yet, signing an alliance with an unreliable partner is at odds with both of these rationales. This
should be particularly concerning for peace scholars and policymakers, since the uncertainty generated by unreliable
partners may increase system-wide conflict. This article provides an answer to this puzzle by arguing that states
continue to form alliances with unreliable partners because they can adopt rational portfolio-diversification strategies.
Drawing on well-developed models from portfolio theory, we present evidence that states design their overall alliance
portfolios to minimize the risks posed by allies with a reputation for being unreliable. Specifically, we show that
unreliable allies are more likely to be pooled into multilateral alliances that dilute risk rather than bilateral alliances,
and that states allied with unreliable partners form a greater number of alliances to hedge against the added risk of
default. Together, our results demonstrate why unreliable partners may not lead to increased conflict initiation, while
also providing a novel explanation for previously unexplained variation in the structure of alliance portfolios. The
article contributes to the literatures on international reputation and the rational design of international institutions
by demonstrating how international reputation matters in subtle and often overlooked ways.
Keywords
alliances, collective security, international reputation, international security, rational design of international
institutions, signaling
Introduction
Why do states ever form military alliances with unreli-
able partners? States largely sign offensive and defensive
military alliances to increase their fighting capabilities in
the event of a conflict (Lake, 1999). In the case of defen-
sive alliances, states may also hope that this increased
fighting capacity will deter aggressors from initiating dis-
putes in the first place (Johnson & Leeds, 2011; Leeds,
2003; Mehta & Narang, 2018). Yet, signing an alliance
with an unreliable ally is at odds with both of these
rationales. First, unreliable alliance partners are, by
definition, demonstrably less likely to honor their com-
mitment in the event of a war. This risk of abandonment
may nullify the benefit of an ally’s additional fighting
capabilities (Snyder, 1984). Second, alliances with unre-
liable partners may even leave a state more vulnerable
than it would have been without an alliance. This is
because states often rely on the added capabilities of allies
to reduce their own indigenous military expenditures
Corresponding author:
narangn@gmail.com
Journal of Peace Research
2019, Vol. 56(3) 379–394
ªThe Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0022343318808844
journals.sagepub.com/home/jpr
(Lake, 1999; Morrow, 1993; Narang, 2018a). Such
burden-sharing can be highly advantageous if a partner
actually honors the alliance, but disastrous if it does not
(Lake, 2009). Third, unreliable allies expose a state to all
of the risk of entrapment from ‘entangling relations’, but
reduce the potential benefits due to a higher risk of
abandonment (Beckley, 2015; Lake, 1999; Narang &
Mehta, forthcoming). Finally, allying with an unreliable
partner may increase uncertainty about whether the
unreliable partner will honor its commitment. Such
uncertainty can lead to conflicting expectations about
the strength of an alliance and increase the chance that
deterrence will fail (Sagan, 1986).
These risks combine to make alliances with unreliable
alliance partners particularly concerning for peace scho-
lars and international policymakers alike. Beyond the
consequences for individual states, unreliable allies serve
to increase uncertainty about which acts of aggression
will trigger war (Sagan, 1986). This uncertainty, in turn,
may increase system-wide conflict (Fearon, 1995;
LeVeck & Narang, 2017a; Narang, 2014, 2015a). And
yet, despite the risks from allying with unreliable part-
ners, previous research has demonstrated that states con-
tinue to form alliances with such states (Crescenzi et al.,
2012; Gibler, 2008; LeVeck & Narang, 2017a; Mattes,
2012). Therefore, the question remains: given the poten-
tially disastrous consequences that states risk, why do
they continue to form alliance agreements with demon-
strably unreliable partners?
In this article, we argue that states continue to form
alliances with some unreliable allies because they can
adopt rational portfolio-diversification strategies. These
strategies allow states to benefit from the military cap-
abilities of unreliable states while minimizing the risk of
abandonment. Our primary theoretical innovation is to
leverage well-developed models from portfolio theory
(Markowitz, 1959). Using the framework of portfolio
analysis, we generate testable propositions about how
states structure their alliance portfolios. Specifically, we
present three pieces of evidence that states design their
overall alliance portfolios to minimize the risks posed by
allies with a reputation for being unreliable. First, we
show that unreliable allies are more likely to be pooled
into multilateral alliances that distribute risk better than
bilateral alliances. Second, we further show that states
internally construct these multilateral alliances to reduce
the trade-off between additional capabilities and addi-
tional risk of abandonment. Third, we demonstrate that
states in bilateral alliances with unreliable partners also
hedge against the risk of abandonment by signing a
greater number of bilateral alliances.
Taken together, our results serve to demonstrate
how remarkably general the Nobel prize-winning logic
of portfolio theory is across domains, while also con-
tributing to the literatures on international reputation
and the rational design of international institutions
(Koremenos, Lipson & Snidal, 2001; Mattes, 2012).
More importantly, our theory and evidence provide an
explanation for why unreliable partners may (happily)
not lead to a large increase in dispute initiation
throughout the international system. Since the struc-
ture of alliance portfolios is visible to other states in the
international system, portfolio-diversification strategies
that reduce the risk of abandonment should also
increase their deterrent effect.
In addition to explaining the puzzle of unreliable
allies, a second important motivation and contribution
of this project is to provide a novel explanation for the
structure of alliance portfolios. To date, variation in alli-
ance portfolios has generally been used as an explanatory
independent variable, rather than the phenomenon to be
explained (Bennett & Stam, 2000, 2003). For decades,
scholars of international relations have exploited varia-
tion across alliance portfolios as a conceptual proxy for
the overall foreign policy similarity. Since Bueno de Mes-
quita (1975), researchers have used either Kendall’s
b
score or Signorino & Ritter’s (1999) S-score to identify
alliance clusters and to measure the extent to which those
clusters overlapped or were discrete (Bueno de Mesquita,
1978, 1981, 1988; Ostrom & Aldrich, 1978; Stoll,
1984; Stoll & Champion, 1985). Alliance portfolios
were thus treated as revealed preferences over security
issues, as states with greater commonality in their alliance
portfolios were generally assumed to have more similar
security interests (Altfeld & Bueno de Mesquita, 1979).
Although useful, here we are the first to demonstrate that
variation in alliance portfolios is not purely driven by
foreign policy similarity, but that it also reflects attempts
to mitigate the risk of abandonment in the event of war.
Literature: The role of reputation in alliances
Previous work has argued that alliance agreements can be
usefully analogized to bonds (Conybeare, 1992; LeVeck
& Narang, 2017a; Mattes, 2012). Similar to how loan
applicants seek loans from potential lenders in order to
increase their financial security, states in the interna-
tional system seek to negotiate alliance contracts in order
to increase their national security. And just as lenders
must ultimately issue loans based only on the expectation
that a recipient will eventually repay their debt, states in
the international system must enter alliance contracts
380 journal of PEACE RESEARCH 56(3)
To continue reading
Request your trial