This article examines the impact of loss-aversion, as defined by prospect theory, upon the responses of the American public and Congress to U.S. military intervention abroad. In particular, it asks whether the manner in which the rationale for the intervention is framed (whether as intended to avert a loss or secure a gain) will influence the amount of support mustered for the intervention. Justifications offered for the active use of U.S. force abroad are content-analyzed to determine whether the main justification stresses avoiding a loss to an acquired geopolitical position (a "protective" intervention) or securing a net foreign policy gain (a "promotive" intervention). The extent of shifts in presidential popularity occasioned by the intervention are compared, controlling for the rationale offered and the effort made by the president (in the form of relevant speech-making) to justify the intervention. Regression analysis indicates that, with these controlling influences, a protective intervention is expected to elicit a six percentage point approval differential over a promotive intervention. The article also examines the difference that the two sorts of justification make with regard to the likelihood that Congress would pass a resolution supporting the intervention. Although the evidence is more ambiguous here, it indicates that a protectively framed intervention comes to being a necessary condition for a supportive resolution, though it is not a sufficient condition.
A better understanding of the domestic foundations of external affairs requires a firmer grasp of the national setting within which foreign policy is imbedded-particularly the play of relevant domestic preferences and the associated concerns of political leaders. In this regard, some light has been shed on the link between electoral calculations and foreign policy decisions (e.g., Stoll 1984; Gaubatz 1991; Nincic 1991), as well as on the interaction of foreign policy and popular opinion (e.g., Monroe 1979; Page and Shapiro 1983; Russett 1989). To some extent, too, the domestic drives to which decision makers must respond have been addressed, especially the bureaucratic influences (Halperin 1974; Hilsman 1990), segmental economic pressures (Cohen 1994), and domestic instabilities (Levy 1989; Morgan and Bickers 1992; Ostrom and Job 1985) that foreign policy must sometimes remedy More recently still, some attention has been paid to the impact of domestically held ideas on the conduct of foreign affairs (e.g., Goldstein and Keohane 1993).
However, we continue to know little about the impact of those behavioral and perceptual predispositions that are not directly linked to parochial interests or structured beliefs but that, nevertheless, shape the domestic political reward structure by which foreign policy decisions are guided. Rooted in social psychology, these predilections affect the responses of entities at various levels of social aggregation to international events, as well as their feelings regarding different tools of foreign policy Such inclinations cut across a variety of interests and tend to be shared by those with differently constituted ideas about international affairs. Accordingly, they may be of more value in explaining systematic inclinations in foreign policy than the variations it exhibits.
The aim of the research presented here is to examine one cognitive predilection that, along with the political interests of top decision-makers (the president in particular), may influence foreign policy choices. Specifically, I will ask whether the political system's response to the use of military force abroad is shaped by a loss aversion characterizing both the public and Congress, possibly imparting a status-quo bias to foreign policy or, at least, to the manner in which major foreign policy decisions are framed.
LOSS AVERSION AND ITS FOREIGN POLICY IMPLICATIONS
The consequences of loss aversion have been most extensively examined within the context of prospect theory (Kahneman and Tversky 1979, 1984; Thaler 1980), a body of thought challenging core assumptions of expected utility theory and whose predictions have enjoyed substantial experimental confirmation (Knetsch and Sinden 1984; Quattrone and Tversky 1988; Knetsch 1989, see also Levy 1992). …