Reputation, History, and War
| Author | Stephen B. Long,Jacob D. Kathman,Mark J.C. Crescenzi |
| Published date | 01 November 2007 |
| Date | 01 November 2007 |
| DOI | http://doi.org/10.1177/0022343307082048 |
| Subject Matter | Articles |
651
Introduction
The role of historical violence between nations
in the incidence of militarized conflict is well
established. The entire rivalry literature, for
example, is built upon the premise that politics
among nations should be studied as long
temporal relationships, where multiple con-
flicts cannot be considered independent from
one another (see Diehl & Goertz, 2000).1
More generally, Crescenzi & Enterline (2001)
demonstrate that a history of violence between
nations is a powerful predictor of future mili-
tarized disputes. The argument is fairly intu-
itive: interstate conflict leaves an indelible
mark on the nations involved, and these
© 2007 Journal of Peace Research,
vol. 44, no. 6, 2007, pp. 651–667
Sage Publications (Los Angeles, London, New Delhi
and Singapore) http://jpr.sagepub.com
DOI 10.1177/0022343307082048
1For a description of the rivalry literature and an explana-
tion of the differences between existing measures and our
approach, see Crescenzi & Enterline (2001: 410–412). For
a broader review of the rivalry literature, see Diehl &
Goertz (2000) and Colaresi & Thompson (2002).
* Author names are listed alphabetically. Data assembly was
conducted with the help of EUGene 3.03 (Bennett & Stam,
2000). Stata 9.2 was used to generate the statistical results
presented in this article. This research was conducted
with the support of NSF grant #SES-0450111. The data
used in this article are available at http://www.prio.
no/jpr/datasets or upon request forom the authors. Please
direct correspondence to crescenzi@unc.edu.
Reputation, History, and War*
MARK J. C. CRESCENZI
Department of Political Science, University of Mississippi
JACOB D. KATHMAN
Department of Political Science, University of North Carolina
STEPHEN B. LONG
Department of Political Science, Kansas State University
This article investigates the role of direct and reputational information in the onset of interstate war.
Scholars have recently identified the importance of separating the phenomenon of conflict from the
rare event of war. Building on earlier work concerning the role of reputation and history in the onset
of militarized interstate disputes, this article argues that states in crises face competing pressures brought
on by their history of interactions with their opponents and their opponents’ reputations generated
through interactions with other states. While historical conflict reveals private information regarding
the credibility of state demands, this history also generates constraints upon the ability of governments
to seek peaceful resolutions to the current crisis. An empirical analysis supports the hypothesis that both
a direct history of conflict within the dyad and reputational histories for conflict increase the likelihood
of war onset. These results hold for a sample including all dyads 1817–2000 and a sample including
politically relevant dyads in the same period. The results also suggest that contiguous states are more
likely to go to war with each other, as are pairs of major powers, while democracies and pairs of minor
powers are less likely to go to war with each other. These results support previous findings on the influ-
ence of these factors on the likelihood of war onset.
memories influence foreign policy decision-
making in times of crisis. Further, recent
research indicates that indirect historical vio-
lence may also impact the likelihood of inter-
state conflict (Crescenzi, 2007). Together,
these two streams of information place crises
and crisis behavior in macro-relational and his-
torical contexts.
In this article, we focus on the question of
whether these contextual sources remain
important when states decide to go to war.
Historical and reputation contexts may be rel-
evant when political crises turn violent, but this
does not necessarily mean that these contexts
are also relevant when states consider taking the
more severe step of going to war. Do these
factors permanently inhibit the ability of states
to negotiate their way out of crises, or does the
specter of war induce caution?
Our expectation is that the historical and
reputation contexts have an important
inflammatory impact on the decision to go
to war. That is, when states engage in crisis
behavior within a context rich with direct
and indirect historical conflict, the onset of
war is more likely. We test this hypothesis
using semi-parametric hazard analysis on
data for war onset covering all international
dyads from 1817–2000, as well as a sample
containing only politically relevant dyads
(Maoz & Russett, 1993). Our results indi-
cate that a direct history of conflict within a
dyad increases the odds of war onset consid-
erably. Additionally, the odds of war onset
increase when either state within the dyad
has a reputation for being conflict-prone.
Before we endeavor to convince the reader
of these findings, we begin in the next section
with a brief discussion of the key concepts of
this study: war, the direct historical relation-
ship, and reputation. We then provide a
theoretical discussion of the causal ties
between behavioral context and war, in order
to produce specific, testable hypotheses. Our
research design section addresses the use of
semi-parametric hazard models to conduct
the analysis, as well as some issues surround-
ing the use of models that compensate for
selection effects (in this case, the selection
effect of entering into a dispute). The analy-
sis follows, along with a brief discussion of the
results and remaining research tasks.
Concepts
There are three concepts that are fundamental
to this research. The first is war, the phenom-
enon for which we seek a partial explanation.
Most readers have an intuitive understanding
of this rarest form of conflict, and the modern
discourse on the phenomenon of war is mature
and relatively uncontroversial (Wright, 1965;
Small & Singer, 1982). Here, we are concerned
with the factors that lead states to engage in
wars generally, and we distinguish between war
and the broader concepts of conflict, such as
disputes and crises. Interstate war is a label
reserved for historical events between at least
two nations where militarized violence is sus-
tained and severe enough to generate signifi-
cant casualties.2We do not address the
characteristics of severity, intensity, or duration
of war in this article, although we hope to do
so in future work.
The second concept requiring explanation is
the notion of a direct, dyadic behavioral history.
Here we borrow directly from Crescenzi &
Enterline’s (2001) conceptual dynamic model
of interstate interaction. This model defines the
behavioral historical relationship between two
countries as an evolution of information involv-
ing change motivated by shocks and decay.
Interstate interactions inform the historical
relationship in both conflictual and cooperative
directions, and the lack of interaction degrades
this information over time. In the absence of
activity, the behavioral history between two
states is defined as neutrality. The occurrence of
conflict shocks this relationship negatively, but
journal of PEACE RESEARCH volume 44 / number 6 / november 2007
652
2The casualty threshold we use to identify wars, as distinct
from other militarized disputes, is 1,000 battle-deaths
(Small & Singer, 1982).
To continue reading
Request your trial