Graphical analyses of international relations: On the determinants of militarized conflicts

Abstract

This article argues for the importance of graphical analyses examining the patterns of data available in the studies of international relations. It posits that there has been a relative lack of graphical analyses in the discipline of international relations, compared to other disciplines such as sociology. We could learn a significant amount of information from the well-performed graphical analyses, which are compatible to and produce important patterns that are not easily revealed from econometric or tabular analyses. In this article, the importance of graphical analyses is evidenced by examining the determinants of the onset of militarized conflicts, one of the most researched topics in the field of international relations.

Acknowledgments

I would like to thank the editor N. Prabha Unnithan and the anonymous reviewers for helpful suggestions that improved the article.

Notes

1 The use of Maoz's modified version instead of the COW original data is based on the finding (Maoz, 2001a) that the COW data set contains dispute cases where states on one side of a multilateral contest may never have threatened, displayed, or used force against states on the opposing side (e.g., Bulgaria and Japan on the opposite sides in World War I). Maoz corrects this and other problems to produce more accurate militarized interstate dispute data (see also Oneal et al., 2003, 376).

2 The two data sets used in this article are not the only international conflict data available. There are other data sets that have been used to examine the causes of international conflicts (see Leng & Singer, 1988, for a general discussion on the different types of interstate conflicts such as interstate dispute, militarized interstate dispute, militarized interstate crisis, and interstate war, as well as the relationships among them). For the descriptions and discussions on those other data collections, see the SIPRI website at http://www.sipri.org/contents/conflict/conflictdatasets.html and Rousseau (2005, esp., 100–105). The SIPRI website provides brief descriptions of 16 different data collection projects on international conflicts. Rousseau (2005) discusses some other data collection efforts on international conflicts such as Huth (1996) and Huth and Allee (2002) on territorial conflicts and Sherrman (1994, SHERFACS).

3 The conflict onset case is defined as “connected” in the network if it has a connection to any other conflict onset cases in the network. The conflict onset case is defined as “isolated” or “disconnected” in the network if it does not have any connection with other conflict onset cases in the network.

4 The measurements and data sources in this part are from Russett and Oneal (2001) and Oneal and Russett (2005).

5 The monadic conflict onset represents a dispute onset of each state, in contrast to the dyadic conflict onset, which represents a dispute onset of each dyad. This distinction is from the level-of-analysis discussions in international relations where monadic analysis focuses on the state, dyadic analysis focuses on the dyad, and system analysis focuses on the system.

6 There have been other efforts to identify global maps of international conflicts (i.e., the monadic patterns of international conflict). For example, the Nobel Foundation provides global maps of 20th wars (1899–2001, interstate, colonial and civil wars) and their casualties (http://nobelprize.org/educational_games/peace/conflictmap/). ViewConflicts (Software for Visualizing Spatiotemporal Data on Armed Conflicts) provides global maps of arms conflicts (1946–2004, internal and internationalized conflicts) (http://www.svt.ntnu.no/geo/forskning/konflikt/viewConflicts/).

7 Brandes et al. (1999) posit that “[given] the fact that already a simple description of the data in the form of a matrix is difficult to read, it seems obvious that an exploration of the data through tables becomes practically impossible. In contrast, a visual presentation allows basic features of the network, as well as a great number of additional information on its structural characteristics, to be observed” (77).

Additional information

Notes on contributors

Hyung Min Kim

Hyung Min Kim received his B.A. in political science from Sogang University, Seoul, Korea, M.A. in political science from the State University of New York-Buffalo, and Ph.D. in political science from the University of North Carolina-Chapel Hill. He has published in such journals as Armed Forces and Society, Defence and Peace Economics, International Political Science Review, American Sociological Review, Peace Economics, Peace Science, and Public Policy, Review of International Political Economy, and Journal of Peace Research. His research interests are international relations theories, security studies, East Asian international relations, and research methods (including social network analysis).