a1 Social Science Research Council/MacArthur Foundation Fellow in International Peace and Security and a Ph.D. candidate in the Department of Political Science, Columbia University, New York.
a2 Associate Professor of Political Science at the Institute for War and Peace Studies, Columbia University, New York.
Contemporary balance-of-power theory has become too parsimonious to yield determinate predictions about state alliance strategies in multipolarity. Kenneth Waltz's theory predicts only that multipolarity predisposes states to either of two opposite errors, which this article characterizes as chain-ganging and buck-passing. To predict which of these two policies will prevail, it is necessary to complicate Waltz's theory by adding a variable from Robert Jervis's theory of the security dilemma: the variable of whether offense or defense is perceived to have the advantage. At least under the checkerboard geographical conditions in Europe before World Wars I and II, perceived offensive advantage bred unconditional alliances, whereas perceived defensive advantage bred free riding on the balancing efforts of others.