Leader Tenure, Domestic Constraints, and International Conflict

Open Access
Bak, Daehee
Graduate Program:
Political Science
Doctor of Philosophy
Document Type:
Date of Defense:
February 18, 2013
Committee Members:
  • Glenn Hunter Palmer, Dissertation Advisor
  • Glenn Hunter Palmer, Committee Chair
  • David Scott Bennett Jr., Committee Member
  • Christopher Jon Zorn, Committee Member
  • Joseph Wright, Committee Member
  • Alessandro Bonanno, Special Member
  • Leader
  • Tenure
  • Domestic Constraints
  • International Conflict
  • Predation
  • Time Horizon
  • Game Theory
My dissertation provides a theory about why political leaders pursue reckless and adventurous foreign policies at some point of tenure while trying to avoid contentious and costly foreign policy in other times during their tenure. I investigate how a leader’s decision to initiate a costly military conflict is constrained by domestic audiences, how the level of domestic constraints systematically changes over the course of tenure, and how the changing pattern differs between democracies and dictatorships. More specifically, I argue that leaders are highly constrained to initiate a costly conflict when the domestic audience's replacement attempt is credible ex ante, i.e., when the replacement attempt is less costly and more likely to succeed. Early in the tenure a democratic leader's political survival is well protected by institutions and domestic consent, so that a replacement attempt is more costly and less likely to succeed than later in the election period; dictators, however, are vulnerable to violent domestic challenges before consolidating dictatorial power, so that a replacement attempt is less costly and more likely to succeed early than later. I hypothesize that the likelihood of conflict initiation decreases in a democratic leader's tenure but increases in a dictator's tenure. I conduct a series of empirical tests and find strong empirical evidence for the hypotheses. My dissertation also examines whether a similar theoretical logic can be applied to only democracies. More specifically, I examine how a democratic leader's sensitivity to potential electoral punishment changes over the course of tenure. I argue that democratic leaders are highly sensitive to domestic electoral punishment when they have strong electoral incentives and when the electoral outcome is highly uncertain. I demonstrate in a game-theoretic model that a leader’s sensitivity systematically affects his or her conflict behavior through audience costs and the costs of war. Equilibrium concepts show that highly sensitive leaders are less likely to initiate a military conflict than less sensitive leaders. Using a fine-grained monthly dyadic data set on the U.S. presidents between 1953 and 2001, I find that presidents tend to be less likely to initiate a military conflict when their personal electoral fate is at stake and the electoral outcome is highly uncertain, and that presidents in the second term are more likely to initiate a conflict than those in the first term. Further, I turn my attention to authoritarian regimes. In particular, this research examines how authoritarian time horizon, i.e., how long an authoritarian leader expects to stay in office, affects an authoritarian leader’s conflict behavior. I develop a simple formal model of autocratic leaders’ predatory behavior in the shadow of domestic constraints in terms of the risk of leadership failure and post-tenure punishment. The model suggests that the likelihood of unsafe or irregular leadership failure be negatively related to the level of domestic predation, which challenges and supplements the conventional Olsonian perspective. Assuming that a military conflict initiation is a type of autocrats’ predatory behavior that enriches their private goods, I run empirical tests on the relationship between the likelihood of different types of leadership failure and the number of militarized dispute initiations. Two-stage models with nonparametric bootstrapping show significant empirical support for my hypotheses, implying that when autocrats are concerned about unexpected and violent domestic challenges, they are highly constrained to use force externally. The overall theme of my dissertation is the relationship between a leader’s tenure and the propensity of international conflict initiation. The major implications of the theoretical arguments and empirical findings are: a) leader is a unit of analysis whose foreign-policy interests and decision-making power systematically change over the course of tenure; b) a threat to leadership survival is a major source of domestic constraints rather than a diversionary motive; and c) democratic leaders are not always more constrained by domestic audiences in their foreign policy than dictators.