argued to produce variations in conflict behavior not attributable merely to state-level factors.
1
This section briefly discusses the dominant empirical findings and theoretical limitations of each
approach.
The political survival approach argues that the risk of losing office, conditioned by domestic
institutions, affects the utility of foreign policy alternatives (i.e., Bueno de Mesquita et al. 2003),
but empirical work inspired by it leaves important questions unanswered. Gelpi and Grieco (2001),
for example, argue that new leaders are likely to be targeted in crises because opponents take
advantage of their initial domestic weakness and an associated willingness to make concessions.
This explanation is at odds with further research on the question, however, as Chiozza and Goemans
find that new leaders both attract trouble, as they are targeted in crises more often (2004b), and
look for it, initiating crises more often than more experienced leaders (2003). Explanations for each
of the latter findings hinge on leaders bargaining “harder” at some point in their tenure, as a result
of either early security (2003) or late-term vulnerability (2004b), and the question of which leaders,
new or experienced, bargain “harder” remains open. Further, neither study makes a compelling
case for why the risks of losing office across regimes should be a source of uncertainty. If it is not,
then it is not clear why inefficient outcomes like conflict should occur even when leaders are known
to be vulnerable at home.
The personal characteristics approach suggests that leaders may be able to develop their own
reputations for resolve (Chiozza and Choi 2003; Huth and Allee 2002), but it lacks an underlying
theory to inform empirical tests and make sense of apparently conflicting results. Huth and Allee
(2002) find that new leaders in democracies tend to both make and win concessions in territorial
1
Gaubatz (1991) also finds a relationship between election cycles and wars in democracies and
makes a political accountability argument based on cycles of state-society relations. The basic
tenure-conflict result, however, has since been shown to apply to both democracies and autocracies.
3
Convention