5
Lawrence Jacobs and Robert Shapiro (1999), based on an intensive examination of Lyndon Johnson’s
foreign policy and polling in 1964 and 1965, suggest that presidents become more responsive to public
preferences during election years (but attempt to mold public opinion to their liking by leading in non-
election years). In all, these findings suggest that leaders are highly responsive to public preferences
when they make foreign policy during election years.
Principal-agent models rely on the assumption of asymmetric information and potentially divergent
goals between actors. The principal (the public) contracts with an agent (the elected leadership) to
perform a certain task (govern in the public's interests). Yet, leaders have a great deal more information
about the situation and the nation's interests when it comes to any specific issue. Since the public cannot
evaluate the quality of the leader's decisions directly, it has to rely on another measure, such as the
success or failure of a policy, in order to determine whether the leader is "good" or "bad" (Morrow, 1991;
Richards et al., 1993; Downs and Rocke, 1994; Smith, 1996). This situation then creates an incentive for
leaders who are facing difficult elections to become involved in a successful international conflict in order
to cause the public to perceive them as "good" leaders (Hess and Orphanides, 1995). By the same token,
if a country is involved in a conflict that is going poorly, leaders might either attempt to end it
successfully through escalation or temporarily stave off a loss in the hopes that they will later pull off a
dramatic success to revive their electoral prospects ("gambling for resurrection") (Downs and Rocke,
1995). While an incentive exists to use force to help an election run, voters will be suspicious of leaders’
actions when they are very far behind an opponent or if the foreign action appears "too easy" (Richards et
al., 1993). These findings imply that an inverted "U" should describe the influence of elections on foreign
policy: little effect if a leader is far behind or far ahead; more foreign policy activity should occur when
an election is closely contested.
Finally, elections might have no influence on foreign policy behavior. Realist theory suggests that
foreign policy is more responsive to international considerations such as the power balance and anarchy
(Morgenthau and Thompson, 1985) and that election years are no different than other years since
domestic factors have little influence on foreign policy at any time. Empirically, some researchers have
Convention