1
Introduction
Theories of international relations frequently assume, explicitly or implicitly, that states and na-
tional leaders are interchangeable as both theoretical and empirical units of analysis, and researchers
often test state-level theories on the behavior of leaders and vice versa (Chiozza and Goemans
2004a). This is particularly common for theories of reputation, in which leaders, not states, are
seen as chiefly responsible for developing and maintaining reputations; Schelling’s (1966) discus-
sion of reputation-building by Kennedy, Khrushchev, and Mossadegeh is a prime example. Such
an assumption also appears in psychological theories of conflict (i.e., Jervis 1976) and in domestic-
political theories in which leaders shape foreign policy in response to threats to their political
survival (Bueno de Mesquita et al. 2003, Clark and Nordstrom 2005).
However, when the preferences and incentives of leaders diverge from the nominal interests of
the state, perhaps as they maximize their chances of political survival rather than the national
welfare (Bueno de Mesquita et al. 2003; Goemans 2000), this may be a particularly dangerous
assumption. On questions ranging from diversionary war (Chiozza and Goemans 2003, 2004b) to
the initiation and termination of conflict (Bueno de Mesquita et al. 2003; Clark et al. 2005; Gelpi
and Grieco 2001; Goemans 2000; Horowitz et al. 2005) to choices of bargaining strategies (Chiozza
and Choi 2003; Huth and Allee 2002) and the development of reputations (Guisinger and Smith
2002), the incentives and characteristics of national leaders appear to answer questions on which
traditional state-based perspectives are silent.
As national leaders become the primary units of analysis in empirical studies, they nonetheless
remain critically underdeveloped as theoretical units of analysis. This paper seeks to correct this
imbalance by developing a formal theory of national leaders in international politics that better
captures relevant leader characteristics like resolve and reputation and more accurately character-
izes the strategic nature of international interactions. Adapting a general model of the political
1
Convention