There are two key concepts in this paper: coalitions and abandonment. There is already a healthy debate on how to define a coalition; however, the concept of abandonment (or defection) has not received the same treatment. While a seemingly straightforward concept, abandonment can be difficult to operationalize.
When studying coalition defection, Weisiger (2016) defines coalition abandonment “as any case in which a country ceases to engage in organized military efforts against the enemy contrary to the wishes of its coalition partners.” This definition combines a behavioral attribute (not fighting) on the part of one state and a policy preference (keep fighting) from another state to code abandonment. While it seems clear that ceasing to fight while some of your coalition partners are still fighting is one way to look at abandonment, it does create potential inconsistencies and false positives. The two cases that come to mind are 1) a form of coalition entrapment and 2) a state with limited objectives.
The first case is a variant of entrapment from the alliance literature; however, here a state has to keep fighting a war it has already joined or be considered a defector. Take the coalition that formed during the 1991 Gulf War. The members of the coalition joined to achieve very clear objectives outlined in various Security Council resolutions, most directly the removal of Iraqi forces from Kuwait. However, had President George Bush decided to march on Baghdad to remove Saddam Hussein, any member that did not join the United States on the road to Baghdad would be considered as having abandoned the coalition. In this example, the coalition’s objectives expanded after the war started, something that we know can happen from the war bargaining literature, and potentially not all members agreed with the new objectives and ceased operations after achieving the original goals. This does not seem like abandonment, at least not in the same way as exiting a war because you made a separate peace and leaving your once partners to fend for themselves.
For the second case, imagine a state joining a war with the expressed purpose of taking a piece of territory; say a colonial holding of one of the belligerents. Here the state joins the war with a specific and narrow objective that is isolated from the larger conflict. After taking the territory, the state may elect to withdraw from the fighting even though the rest of the coalition keeps fighting. Under the current definition of abandonment this would appear as a successful wedge strategy that split a state from the coalition when in reality the state lived up to its commitment. Again, this does not feel like abandonment since the state in question did exactly what it said it would do.
The case I am making clearly draws from a similar debate that took place inside the alliance reliability literature. Early studies relied upon a behavior, fighting with an ally, to determine if states honored their alliance commitments. The results from these studies gave the impression that states rarely honored their commitments. It was not until the creation of the Alliance Treaty Obligations and Provisions (ATOP) project, and its focus on the exact commitments made in a treaty, that we learned that states tend to take their commitments seriously. What we also learned is that states very carefully define their obligations in order to meet their objectives. Given this, it seems justified to assume that states join coalitions to obtain certain objectives and they will tailor their commitments to the coalition to match their goals.
Following the ATOP example, in order to study the concept of coalition abandonment we need to carefully examine the commitments individual states made to their coalition partners. Then we can compare the commitments to the observed behavior in order to judge when a state abandoned a coalition. This is a large undertaking and one handicapped by the fact that coalitions do not require any sort of formal written document (although many are extensions of pre-war alliances or are covered by war time treaties). However, this seems like the natural evolution given the clear policy relevance an understanding of coalition abandonment could have on future efforts in building international military coalitions.