Rost and Schneider: A Global Risk Assessment Model for Civil War Onsets

N Rost and G Schneider, “A Global Risk Assessment Model for Civil War Onsets.”

The authors create a model for the risk that different countries face of civil war over the next 5 years using multivariate logit models. Their goal is to establish an effective early warning system that will allow IOs to more effective allocate scarce resources to mitigate potential conflict eruptions.

Early warning systems typically use two types of date: “events data” or “standards-based data”. Events data models typically use news feeds gathered and look at conflict in a specific region and can predict the onset of conflict in a very short period of time. Standards-data use data that is gathered over a much longer period of time and has an annual track-record upon which to build.

The authors argue that, while state strength has been highlighted as a key determinant of the stability of states, one factor that is not added to the equation is the state’s enforcement of basic human rights. While state strength may lead to instability, if a state is actively discriminating against a population, it is expected to potentially exacerbate the problem.

The authors go through a variety of explanations for the onset of civil war, from political to economic to demographic.

They then construct their model. They find that their human rights variable is highly statistically significant with the onset of civil conflict, among a variety of other important variables: civil war decreases with increasing economic development; oil exporters are at higher risk; political instability and mountains do not correlate; population size has no significance; democracy correlates with civil war; military regimes experience more civil war.