Choosing Sides: Testing a Political Proximity Model

Abstract

I seek to explain and predict how great powers choose sides in others' conflicts by examining Soviet support behavior in 403 diverse conflicts during 1950-87. Using alliance and ideology indicators, I constructed a political proximity model of how choices will be made. For interstate conflicts, support for the politically closer belligerent is expected. As for civil conflicts, support for the embattled government at the expense of insurgents is the expected norm. The model correctly anticipated Soviet support behavior in 71% of the cases; it held up well across the 38-year period studied. My findings demonstrate the usefulness of alliance-based concepts of expected utility as well as the limitations of a simple action-reaction perspective in accounting for choosing sides. They also indicate the primacy of bipolar rivalry, the inadequacy of Cold War views of Moscow as the bastion of world revolution, and the limited impact of bureaucratic factional politics and broad changes in international and domestic politics on how Moscow chooses sides.