Today (today being, uh, long ago), I posted the first draft of a new working paper, “Interstate Coalitions and Strategic Restraint” ( find it on the Research page). It’s the second paper in my broader project on coercion through coalitions, and this one—despite the absence of such references in the paper—came out of thinking about coalitions as sheriff’s posses: they collect a lot of power under one banner, but if they don’t break up after bringing in the first fugitive, they can be a threat to the order they’re nominally out to preserve. The posse analogy is surely not new to the field, but it seems to me that using it to think about the problem of strategic restraint is, and I use it to model the alignment decision of an observer that watches a coalition in action and worries about facing some or all of its members in a future crisis.
More after the jump…
The game goes something like this. (1) A coalition is fighting a war today against a target. If it wins, the leader and possibly partner will have future a crisis with an observer, (2) The observer can choose to remain neutral, side with the coalition, or side with the target. Intervening is costly, but opposition and support represent different tradeoffs: (a) opposition may help eliminate the coalition today, but if it fails, the observer is weakened for the crisis tomorrow; and (b) support lets the observer strengthen itself for tomorrow by carving up the target with the coalition, but of course it helps the coalition win today. (3) After the war, the coalition disbands or remains intact. Finally, (4) there is a subsequent crisis between the observer and either (a) the coalition’s leader, if it disbanded, or (b) the intact coalition.
For all the simplicity of its setup, the game turns out to be pretty complex in the solution, but there are some clear general implications to be drawn from it. First, coalitions with very similar foreign policy preferences are the most likely to provoke opposition from observers, because they can’t credibly promise to disband after the current crisis. More diverse coalitions, on the other hand, can make those promises more credibly, and they’re less likely to find themselves in larger, expanded wars. A corollary is that acting through formal institutions populated by your friends may not reassure quite as many potential opponents as you’d hope, because an aggregation of power less likely to stick around will encourage fearful observers to save the costs of intervening. Second, when opposition is risky and the observer is resolute, we see something called disingenuous bandwagoning, where the observer actually sides with the coalition in order to capture some of the spoils for use in its own defense tomorrow, which helps it split the coalition. While I think the coalitional diversity point is cool, this is the one that excited me, because I didn’t see it coming until the model’s solution started taking shape, and, if you read the paper, I think it provides a plausible explanation for why Russia switched from rhetorical opposition to practical support for NATO’s aims in Kosovo in 1999. But this is just a teaser, so I’ll leave the why to the paper itself.
All that said, the paper is in its pretty early stages, so it’ll be interesting to see what form it ultimately takes. If it’s anything like the first coalitions paper, it may not look much like itself at all. The big question, of course, was whether to try to jam some empirical tests in there, but there’s just no room. The good news is that my first cut at testing the opposition argument is pretty good, but, alas, it looks like I’ll save the empirics for…a book, maybe?