Explaining War Termination: a Boolean Analysis of Causes
| Published date | 01 January 2003 |
| Author | STEVE CHAN |
| Date | 01 January 2003 |
| DOI | 10.1177/0022343303040001205 |
| Subject Matter | Articles |
49
Introduction
Why are some wars protracted while others
end in short order? Only a few empirical
studies exist to shed light on this question
(for example, those of Bennett & Stam,
1996; Vuchinich & Teachman, 1993;
Werner, 1998). Likewise, there is only a small
body of literature that addresses processes
ending armed hostilities (for example,
Carroll, 1970; Goemans, 2000; Handel,
1978; Klingberg, 1966; Pillar, 1983; Stein,
1975; Wittman, 1979). Thus, research on
warfare has not given nearly as much atten-
tion to why states stop fighting as to why they
fight in the first place. Fox (1970: 2)
remarked over thirty years ago that ‘the
international relations literature is far richer,
except in international law, in its prescrip-
tions for preventing and deterring war than
in its prescriptions for limiting, de-
escalating, and terminating war once it has
started, except insofar as winning inciden-
tally involves termination’. Beer & Mayer
(1986: 95) pointed to the persistence of this
asymmetry when they commented that
‘except for memoirs and historical accounts
of final battles and peace negotiations, it is
difficult to find more than a handful of
general works on war de-escalation and
termination’. More recently, Craig & George
(1995: 228) noted that ‘students of military
strategy have not given much systematic
attention to the problem of terminating
wars, and military planners have also
© 2003 Journal of Peace Research,
vol. 40, no. 1, 2003, pp. 49–66
Sage Publications (London, Thousand Oaks,
CA and New Delhi)
[0022-3433(200301)40:1; 49–66; 030205]
Explaining War Termination: A Boolean Analysis
of Causes*
STEVE CHAN
Department of Political Science, University of Colorado
An understanding of why states fight wars requires an understanding of why they end wars. These
phenomena form interrelated parts of a larger process of interstate bargaining. Yet, compared to the
research on the outbreak of wars, there has been much less attention paid to their termination. These
conflicts end when the belligerents reach convergent expectations that they have more to gain or less
to lose if they stop fighting than if they continue to do so. This article presents several factors hypoth-
esized to hasten or delay this mutual recognition and, therefore, to shorten or prolong a conflict.
Boolean algebra is applied to study the individual and joint effects of these factors on the duration of
23 interstate wars during 1945–92. The results show that short wars stem from different causal combi-
nations, therefore suggesting the inadequacy of any monocausal explanation. Moreover, this analysis
points to the importance of context-specific understanding because the same factors can have different
effects on shortening war and different factors can have similar effects on this outcome under different
circumstances. Contrary to the argument in favor of an incremental escalation of military pressure on
an adversary, massive and intense fighting at the outset of a war has been nearly a necessary though
insufficient condition for a quick ceasefire. This conclusion supports the view that private information
inclines states to fight wars but the revelation of this information in the subsequent fighting puts them
in a better position to reach a settlement.
* The author can be reached at steve.chan@colorado.edu.
21P 03chan (ds) 25/11/02 8:38 am Page 49
traditionally neglected this problem, concen-
trating upon how to start wars and fight them
successfully rather than upon how to end
them if things do not go according to plan’.
Perhaps this phenomenon has to do with
the view that wars end when one belligerent
defeats the other. The stronger side prevails
militarily over the weaker side which is
forced to surrender, thereby bringing an end
to hostilities. This is, of course, not an inac-
curate characterization of some conflicts such
as World War II. Since 1945, however, there
have been relatively few wars that ended with
a clearcut outcome of victory and capitula-
tion. Even for many lopsided military clashes
that fall short of the operational definition of
war (which requires a minimum of 1,000
combat fatalities in the case of the Correlates
of War project), a settlement is reached
before the losing side becomes unable to
resist any further. Therefore, physical exhaus-
tion or military collapse does not explain the
conclusion of most contemporary wars.
States rarely fight to the finish whereby one
combatant completely disarms the other
(Wagner, 2000).
One may seek to understand better the
reasons for war termination by examining
how quickly the belligerents are able to reach
a settlement. The duration of armed conflicts
indicates the ease or difficulty that the
participants have had in developing conver-
gent expectations. Because wars are ineffi-
cient (that is, they entail costs regardless of
their eventual outcome), the parties should
have an incentive to settle their dispute. If
they possessed complete information to
anticipate fully the outcome of their contest,
the combatants would have obviously been
better off if they had negotiated a settlement
without going to war (Fearon, 1995; Werner,
1998). Wars occur (and continue) if one or
both sides possess private information that
inclines them to make different estimates of
the likely outcome. The belligerents must be
jointly optimistic so that their subjective
probabilities of victory sum to be more than
one. Their estimates are neither symmetric
nor zero-sum (Stam, 1996; Wittman, 1979).
Although states routinely engage in verbal
and non-verbal posturing (such as by exag-
gerating their objectives and commitment in
order to bluff their adversaries), the outbreak
of war helps to disclose their true capabilities
and intentions. Unlike bargaining in other
contexts, ‘fighting is a source of information
that is much less subject to manipulation by
adversaries’ (Wagner, 2000: 478). Other
developments during the war, such as battle-
field results and the action (or inaction) of
allies, offer further opportunity to update
estimates and adjust objectives. The revel-
ation of this new information reduces the
parties’ inconsistent expectations, thereby
making an agreement more possible. Because
of the tendency for states to misrepresent,
they must undertake costly communication –
such as the resort to arms – to give credibility
to their signals (Fearon, 1995). Therefore,
war sometimes becomes a necessary prelude
to a negotiated settlement (Blainey, 1988).
The belligerents have an opportunity to
settle at any time during the war (Pillar,
1983; Smith, 1998; Wagner, 2000). They
can extend an offer to their opponent or con-
sider an offer from their opponent such that
its acceptance is more mutually beneficial
than a decision to continue war with its
certain costs and uncertain prospects. War
will go on until both sides believe that a
ceasefire will make them better off than
carrying on with the hostilities (Schneider,
1987: 110). A concern for war duration
accordingly directs attention to those con-
ditions that facilitate or retard this mutual
partisan adjustment.
Plausible Explanations
This analysis is about the cessation of large-
scale combat and not the conclusion of a
formal agreement between the belligerents.
journal of PEACE RESEARCH volume 40 / number 1 / january 2003
50
21P 03chan (ds) 25/11/02 8:38 am Page 50
To continue reading
Request your trial