Defense Pacts and International Trade
| DOI | 10.1177/00223433030405003 |
| Published date | 01 September 2003 |
| Date | 01 September 2003 |
| Author | Andrew G. Long |
| Subject Matter | Journal Article |
537
Introduction
Are military security alliances important
factors in the decisions of governments to
raise or lower tariffs, institute non-tariff
barriers, etc.? Do private actors consider the
political–military ties of their home country
when they establish trading relationships
with foreign counterparts? While hypotheses
developed in the literature on this subject are
compelling, they have not been supported
consistently by empirical analyses. Statistical
models demonstrate that partners in a
military alliance have higher, lower, or
similar trade levels than non-allied countries
(Gowa & Mansfield, 1993; Gowa, 1994;
Mansfield & Bronson, 1997; Morrow,
Siverson & Tabares, 1998, 1999; Mansfield,
Milner & Rosendorff, 2000). In short, a
review of the evidence for the effect (if any)
of alliances on the extent of commercial
exchange between states presents one with an
empirical puzzle.
Isuggest that the mixed findings about
the relationship between alliances and trade
are due, in part, to a gap between con-
ceptualization and operationalization. The
difficulty lies in the selection of an indicator
for military alliance that does not match the
concept employed in the theory. Scholars
writing on this topic build their theories on
aconcept of military alliance that considers
the partners to a treaty obligated to defend
© 2003 Journal of Peace Research,
vol. 40, no. 5, 2003, pp. 537–552
Sage Publications (London, Thousand Oaks, CA
and New Delhi) www.sagepublications.com
[0022-3433(200309)40:5; 537–552; 035591]
Defense Pacts and International Trade*
ANDREW G. LONG
Department of Political Science, Florida State University
Theories of international political influences on trade maintain that shared alliance commitments affect
bilateral commercial exchange. The most prominent of these suggests that trade produces a security
externality threatening states, but that a military alliance alleviates these concerns, leading to greater
trade among allies. Yet past empirical analyses produce inconsistent findings with respect to the effect
of military alliances on bilateral trade levels. This article presents a new argument to explain the incon-
clusive findings in previous studies. The potential increase in military power from efficiency gains
through trade is a positive security externality only when a commitment to defend one’s alliance partner
exists. In essence, cooperative security agreements that require states to give military assistance to each
other if attacked (i.e. defense pacts) will lead to greater trade among their members than agreements in
which the members promise neutrality, non-aggression, or consultation. The analysis distinguishes
alliances that include commitments of defense from those that do not, and uses two different sources
of alliance data to examine the effects of defense pacts on trade levels between major powers from 1885
to 1990. The results demonstrate that defense pacts are associated with higher trade among alliance
members, but that trade between members of non-defense pacts is statistically indistinguishable from
trade between non-allies. This article recognizes that a link between economic and security issues exists
and emphasizes their mutual relationship.
*The author thanks Ashley Leeds, Will Moore, Sara
Mitchell, Dale Smith, Stephen Shellman, Michaela Mattes,
and three anonymous reviewers for helpful comments.
Data for replication can be downloaded at http://www.
ruf.rice.edu/~leeds/atop.html and correspondence can be
directed to aglong@rice.edu.
03 JPR 40-5 Long (JB/D) 23/7/03 8:38 am Page 537
one another. An ‘ally’ functions as a military
partner in the event of war in the logic of
the argument. Yet, when designing empiri-
cal analyses, researchers include a much
broader range of security agreements in the
category of military alliance than necessary
to capture the concept in their argument. I
question the choice in previous research
designs to operationalize alliance as any of
three types of formal agreements (defense
pact, neutrality/non-aggression pact, or
entente) when the argument rests on the
commitment of states to aid one another in
the event of war. The type of obligation
undertaken by the members of the treaty is
important; alliances that contain pledges of
military assistance in the event of attack,
defense pacts, should lead to greater trade
among allies. Choosing an indicator that
corresponds to its concept allows me to shed
some light on the empirical puzzle
surrounding the relationship between
alliances and international trade.
Within two different alliance datasets, I
distinguish alliances that include commit-
ments of military assistance in the event of
attack (defense pacts) from those that do not
(non-defense pacts) and examine trade levels
between major powers from 1885 to 1990 to
test the hypothesis that defense pacts affect
international trade. The evidence is com-
mensurate with my argument: defense pacts
are associated with higher trade among
alliance members, but trade between
members of non-defense pacts is statistically
indistinguishable from trade between non-
allies. This relationship between defensive
allies and international trade is not contin-
gent upon polarity in the international
system, as previous arguments contend, and
holds when using data from the Correlates of
War (COW) or Alliance Treaty Obligations
and Provisions (ATOP) collections.
The findings in this study support the
‘realist’ theory of security externalities
(Gowa, 1994) and substantiate arguments
suggesting that power and security may
affect international trade (e.g. Krasner,
1976). However, the argument also lends
credibility to ‘institutionalist’ approaches
advocating that scholars pay more attention
to how states design their relationships with
one another and the international agree-
ments between them (e.g. Downs, Rocke &
Barsoom, 1996; Fearon, 1998; Lake, 1999).
Moreover, this article provides further
evidence that economic issues and security
issues are related, which suggests that ana-
lyzing their connection is important. While
proponents of the liberal peace hypothesis
suggest that trade deters conflict between
states (e.g. Polachek, 1980; Russett & Oneal,
2001), this study finds that security cooper-
ation influences bilateral trade as well.
The article begins by discussing the
alliance–trade relationship. I examine some
common definitions of alliance to illustrate
the need for careful matching of definition,
concept, and statistical indicator, and review
three arguments for why one should or
should not expect a relationship between
alliances and trade. Next, I report on some
mixed empirical evidence regarding alliances
and international trade. In the third section,
Iargue that previous statistical analyses of
the relationship between alliances and trade
produce inconclusive results because
scholars have tested their theories with an
insufficiently discriminatory operationaliza-
tion of alliance. Equipped with this new
insight, the fourth section describes the
research design necessary to evaluate the
effect of defense pacts on trade. I construct
astatistical model to estimate international
trade levels and test my hypothesis on major
power trade from 1885 to 1990. The initial
analysis supports my expectations, and a
subsequent test with an alternative sample
of alliance data produces a similar result. I
conclude with some remarks on the impli-
cations of these findings and suggestions for
further study.
journal of PEACE RESEARCH volume 40 / number 5 / september 2003
538
03 JPR 40-5 Long (JB/D) 23/7/03 8:38 am Page 538
To continue reading
Request your trial