Modeling the factors shaping the support for the use of military force
Support for the use of military force is not a constant. To draw up and test a theory that explains the total complex of variations in support, the following very different perspectives should be taken into account. In Chapters 3 and 4 we looked at the role of ideological and predispositional variables (such as Atlanticism and general attitudes on the use of force) as correlates of attitudes on the use of military force either in general or in specific cases. In Chapter 5 we have discussed the role of time in dynamically shaping support. Following the model developed in Chapter 2 we now turn our attention specifically to the role of situational and contextual variables.
A variety of factors can be brought under the general heading of situational variables, but for the purposes of this study we singled out a few that seem to be particularly relevant. As explained in Chapter 2, they include the nature and characteristics of the conflict in which the use of force is either contemplated or actually exercised, such as the values and interests at stake, the (perceived likelihood of) success, the leadership displayed and the perceived (international) legitimacy of the military action and, finally and particularly, their costs in terms of casualties. Following the distinctions introduced in Chapter 2, we also distinguish a second, more general contextual level consisting of a triangle of influence factors: 1) the nature of the conflict involved and the kind of military action; 2) factors pertaining to the specific characteristics of the country in question, such as its military traditions, national histories, membership of alliances or its geographical situation, including the national discourses arising out of them; and finally 3) the evolution of time and the phase of the conflict in which support is measured, and its duration. While the latter aspect was the focus of analysis in Chapter 5, the present chapter will focus on the first two.
While all of these determinants can be brought under the heading of situational and contextual variables, they touch rather different aspects and hence there is probably a need for different theories to explain their impact and their importance. As far as the methodology employed to study the situational and contextual variables is concerned, we explore these dimensions primarily by looking at the impact of variations in wording in poll questions pertaining to these aspects as observed in a great variety of polls. We use cross-country comparisons as well as measurements of the evolution of support over time.
This approach is not entirely new. As far as international politics is concerned, the use of differences in question wording has been explored before, for instance, by Eichenberg (2005, 2006) and Jentleson (1992; Jentleson and Britton, 1998). Unlike earlier studies, however, we take an explicit transatlantic perspective in this book, by adding data from other countries in addition to the more frequently used poll data from the United States.