Identifying International Networks: Latent Spaces and Imputation

Most analyses of world politics and studies of national security policies recognize the interdependence among the salient actors across the salient issues. Traditionally, international politics has been defined as the scope and extent of the relations among independent countries, thought to be the most important elements in world politics. This means that actors as well as their actions are strategically interdependent (Signorino 1999). Ignoring the interdependence among these phenomena would appear to be a serious oversight that plague attempts to understand, let alone predict, the course of national security policy and world politics more generally. With very few exceptions, quantitative, systematic studies of international relations and national security have assumed that the major actors and actions that comprise world politics consist of unconnected actions and actors. Game theoretic models are legion, but rarely deal with more than two actors at a time. Some beginning attempts to model the interdependency in international relations have appeared in the literature (Ward and Kirby 1987; Gleditsch and Ward 2001; Gleditsch and Ward 2001; Ward and Gleditsch 2002; Gleditsch 2002; Lofdahl 2002), but as yet network models have yet be widely applied in scholarly or policy work on international politics. This is somewhat surprising, since it is evident at first blush that international politics is about the interdependence that appears around the world. Social network analysis is one technique that has been developed to map and measure the relationships and flows among agents. The nodes in the network are the individuals and groups and the links among them illustrate their interdependencies, both in terms of structure and in terms of the flows of information from one node to others. Since the development of the sociogram (Moreno 1934), sociologists among others have been interested in analyzing the linkages among individuals and groups. An interesting early example is found in the early work of Coleman, Katz and Menzel (1957). Most early theoretical advances were based on graph theory as developed and advanced by Frank Harary and his students and collaborators (Harary 1959; Harary 1969; Harary, Norman and Cartwright 1965). The so-called “Columbia school” worked throughout the 1960s and beyond to further advance the substantive findings in this arena of sociology (White 1963; White, Boorman and Breiger 1976). Broader dissemination of these ideas came much more recently with didactic writings (Knoke and Kuklinski 1982; Scott 1991) as well as early