Social network analysis (SNA) originated in the 1930s. Since the 1970s, it has developed rapidly and been introduced into the study of international relations, facilitating the networked processing of empirical evidence, methodological integration, and the expansion of research approaches.

Providing relationship-based theoretical perspective

Since its introduction into the field of international relations, SNA has evolved through three stages. The first stage, beginning in the late 1990s, focused on describing the role of interpersonal networks in international politics. The second stage, starting in the early 2000s, shifted attention to international relations networks with states as nodes. The third stage, unfolding over the past decade, seeks to develop an international relations theory based on network theory. From this stage onward, SNA in international studies has given rise to four key research themes: networked power, network effects, interactions among networks, and the impact of external shocks on networks.

Theoretical innovation is invariably accompanied by methodological transformation. The contribution of SNA to theoretical innovation in international studies is evident in four respects: it highlights the utility of relationships in international politics, moving beyond the traditional focus on state attributes; it enables a more comprehensive understanding of the complex interactions between individuals and structures; it reveals the relational dimensions of existing data; and it offers new approaches to causal inference, with the potential to improve the predictive accuracy of international relations theory.

Serving as tool for measuring socialized relations

Once adapted, the techniques and core concepts of SNA can be applied to define, measure, and compare relationships among states. Among these, a fundamental analytical technique is graphical representation, which uses network diagrams to visualize research subjects and causal relationships.

According to social network theory, differences in the positions of actors within a relational network determine variations in their behavior and their capacity to influence others. Occupying a favorable position within the network confers influence or social power. First, positional differences reflect variations in influence. Nodes with broader social connections can exert influence more effectively, while their hostility is detected more quickly in adversarial networks. Second, nodes closer to the network’s center possess greater influence. They can receive and transmit information and resources more efficiently, granting them more channels to affect others. Third, critical positions amplify influence. Some nodes, though not central, wield significant power owing to their positional advantages. One type acts as a bridge between the network core and its subgroups, rendering them irreplaceable. Another type, less dependent on any single network, participates in multiple relational networks and thus enjoys greater bargaining power.

Causal inference in SNA primarily concerns network effects—that is, the influence of relational structures on the behavior of nodes, or the logic underlying the formation of connections within networks. Currently, graphical modeling remains the main method for inferring causal mechanisms in SNA. In addition to descriptive graphs that depict empirical realities, exponential random graph models (ERGMs) and directed acyclic graphs (DAGs) are also commonly employed.

It should be noted that, as with other methodological approaches, applying SNA to international studies also presents challenges. These include uncertainty about the role of social power, risks of personifying relationships among states, ambiguity of core concepts, and the inherent complexity of causality. Researchers must therefore carefully assess both the applicability and the limitations of SNA in this context.

Dong Zuozhuang is an associate professor from the Zhou Enlai School of Government at Nankai University.