Dyads and Conflict – Beyond Interstate War

The latest wave in the quantitative study of civil war began in the IR subfield, but it was not dominated by a dyadic design. Instead, a country-year approach yielded studies that largely explained what type of countries are prone to civil war, and it was only in a 2009 special issue of the Journal of Conflict Resolution that the use of dyadic data became the norm. 


The new “dyadic” normal can be challenged on a few fronts, some of which mirror Cranmer and Desmarais’s critiques. First, some wars are multi-party; some have more rebels (or external actors) than others. While qualitative studies have delved more into the complex dynamics between rebel groups, the quantitative literature has yet to do so to the same degree (though some recent work on side-switching of militias is moving in this direction).


A second challenge for the dyadic design centers on whom is in the dyad.  Actors in civil conflicts are not necessarily unitary, and the complexity of the opposition is now being explored on a number of fronts, including its nature (with respect to dimensions and power), causes and consequences. Moreover, there is typically a clear distinction between the “action” and the “actor” in classic studies of international politics that center on things like war, treaties, or trade agreements.  Yet, when we look beyond states, actors are often identified by their action (i.e. rebels rebel, opposition movements protest, non-governmental organizations lobby for specific policy changes). This differs to some degree from Cranmer and Desmarais’s critique because this actor-based challenge is a call to consider the composition and constitution of actors in ways that states are typically not addressed in quantitative conflict studies. Similar issues might arise with study of non-state actors in international relations.


The interdependence critique could also be leveled at dyadic studies of civil war, though perhaps differently than for state-to-state theories of conflict. In a number of instances, states face multiple challengers in what are essentially different civil wars (for example, see India’s multiple low-level wars).  Dyads in conflict with the same state are clearly not totally independent. Even across borders, however, there can be links between actors that challenge the state. 


Though there is some move toward network analysis in studies of civil conflict, this has generally been in the area of conflict patterns amongst non-state actors. The usual method for addressing such issues continues to be clustering standard errors, or more rarely using multilevel modeling. A key challenge for employing more nuanced network analysis is the lack of information about the relationships between actors, or even the identification of a stable set of actors.  

Discuss this Article
There are currently no comments, be the first to post one.
Start the Discussion...
Only registered users may post comments.
ISQ On Twitter