“…(Sartori, 2005) 7 The accuracy and utility of the unitary actor assumption is frequently contested in IR (Allison, 1971;Keohane and Nye, 1977;Milner, 1997;Moravscik, 1997;Fearon, 1998;Hudson, 2005), but what matters for our purposes is that a divide exists in the rationalist literature between theories of resolve with unitary actors (Morrow, 1989;Morgan, 1990;Slantchev, 2005), and those that incorporate domestic politics, whether in the form of audience costs (Fearon, 1994), agent-principal problems (Downs and Rocke, 1995), selectorate theories (Bueno de Mesquita and Siverson, 1995), or two-level games (Putnam, 1988). In fact, much of the literature on the signaling of resolve explicitly requires non-unitary state actors, in that a state's ability to signal its resolve depends on the nature of the relationship between a government and its opposition (Schultz, 1998;Levy and Mabe Jr., 2004), or between the government and the public (Fearon, 1994). military, for example, has significant implications on who is "shouldering the soldiering" (Vasquez, 2005;Horowitz and Levendusky, 2011), and analyses of the effects of local casualties in the Vietnam war (Gartner, Segura, and Wilkening, 1997) suggest that the distribution of the conflict's costs also affects attitudes towards the mission in general. In addition to asymmetries in cost between different groups of the public, there also may be asymmetries between the public and its leaders; Chiozza and Goemans (2004) find that whether war is costly for leaders in terms of reduced time in office depends in large part on the regime type, such that leaders are often insulated from the costs of war imposed on their publics.…”