The research team under Lars-Erik Cederman from ETH-Zürich set out to analyse through which conditions and mechanisms campaigns of armed violence against civilians contribute to the escalation from nonviolent to violent situations and to civil wars.

To that goal, data on the ethnic identity of civilian victims of one-sided violence was collected. This novel database, the ethnic one-sided violence database, is based on the Uppsala one-sided violence database.

One of the most important insights that has emerged so far is that ethnic inequality is one of the main drivers of ethnic state violence against civilians, and that ethnic violence against civilians by armed groups tends to escalate conflicts. In the research paper available HERE, they laid out the following four hypotheses:

1. Ethnic exclusion increases the probability of ethnic targeting by state actors during armed conflict
2. Given high levels of ethnic exclusion, whether a conflict is ethnic or non-ethnic does not increase the probability of ethnic violence against civilians perpetrated by state actors
3. State actors are more likely to target the civilian constituency of insurgent organizations if the latter recruit significantly from excluded ethnic groups
4. Given that insurgent organizations recruit significantly from excluded ethnic groups, whether a conflict is ethnic or non-ethnic does not increase the probability that state actors will target the civilian constituency of these organizations

While no solid conclusion can be drawn on the fourth hypothesis, numbers one to three seem to be supported by the data gathered. However, the authors warn us to take these findings with a grain of salt, as the causal relationships between ethnic claim-making, ethnic exclusion, and ethnic targeting are not yet fully disentangled.

The team is working on several papers interpreting the data they collected for their database. For more information on the project, check out the project outputs on our website.